Keyword Difficulty Showdown, Which Keyword Tool Will Give You The Most Accurate Data?

Gael’s Disclaimer: I just finished editing this 10,000-word / 50-minutes-of-video monster Perrin put together, so I will try to keep this short.

In this post we try to establish what the best keyword tool is for regular publishers like us because we spend thousands of dollars monthly producing content and this content makes or breaks our business.

Google being a large part of our traffic, our ability to pick the right keywords and topics to target determines whether we make a profit or a loss on most of our content.

That is why we decided to go super deep on tools and find what according to us is the best keyword research tool.

I stand 100% by the views Perrin expresses in this post and no company paid us to produce any of this content (although some might offer us some money to take it down).

It’s also important to note that Perrin was associated with Niche Pursuits who created (and still has a stake in) Long Tail Pro (he doesn’t work there anymore).

While he tried to stay objective in the test and didn’t seem to give LTP any preferential treatment, it’s my job as the editor to disclose that to you.

Now, enjoy the post!

Update (June 2016)

A lot has changed in just a couple weeks! Since this post was published, we received responses from two of the companies reviewed here (Moz and Ahrefs), and another company (Long Tail Pro) changed its software significantly.

As you might have guessed, most of the hullabaloo revolves around our analysis of keyword difficulty scores, and we now have a bit more information to work with (although we have less information in some areas, too).

Here’s a quick rundown of the new stuff:

The most important implication of all of this is that we have a better understanding of what the various KD metrics actually mean, which makes it easier to interpret the data we collected and, hopefully, arrive at a better answer to the question “Which keyword research tool is most accurate?”

However, with changes to established software (Long Tail Pro) and other keyword research tools in their infancy (Moz)—and in some cases still in beta (Ahrefs)—it just seems prudent to revisit this post as the tools evolve and as more people get a chance to use them.

That said, I’m not going to include all updates in one section. Instead, I’m going to append updates to the relevant sections of the post, so as you read, you can see both the original information and what’s been changed.

In the last couple of months, it seems like pretty much everyone has released a keyword tool (or at least an updated version of their keyword tools).

SEMRush Keyword Difficulty News
SEMRush released their v1 about 1 year ago
Ahrefs around 1 month ago
Ahrefs around 1 month ago
Announcing Keyword Explorer Moz's New Keyword Research Tool Moz
And just a few days later MOZ released their brand new keyword tool

Of course, keyword research tools have been around forever, but none of the truly big SEO research companies (Moz, Ahrefs, and SEM Rush) jumped on the bandwagon until this year.

Those suites of tools have included keyword research features, but they weren’t built for it.

Because of that, other, more specialized tools (like Long Tail Pro and Market Samurai) have more or less dominated the market.

And the one thing the specialized tools had that the research suites didn’t was a keyword difficulty metric (a working one, anyway) — a concrete metric that gave you an idea of how difficult given keyword was to rank for.

In the last 60 days or so, all of that changed…

Each of the three major research suites — Ahrefs, Moz, and SEM Rush — added new keyword research tools. More importantly, though, all of these tools seem to have been built around a keyword difficulty metric.

In my view, this is pretty big news.

As SEO evolves and the big companies mature, the industrial-level tools are starting to slowly cannibalize the myriad smaller tools that speckle the internet marketing landscape by adopting — and in some cases improving — their functionalities.

The problem, of course, is that small, specialized tools often do a much better job within their narrow expertise than bigger, more general tools.

Typically, this isn’t the case forever, since big companies usually have R&D budgets and can iterate more rapidly; it just means that when the big boys decide to add features, they’re often worse than existing tools at launch.

And that’s the question we are trying to answer in this blog post: how do the new keyword research tools from the big companies stack up to (arguably) the best specialized keyword research tool on the market?

We’re going to do this in two ways…

  1. We’re going to execute actual keyword research campaigns for an imaginary website. I’m going to do that on video as a way of reviewing each keyword tool in a very practical setting. Those are just below
  2. We’re going to run a rather exhaustive experiment to compare the accuracy of the keyword difficulty scores for each tool. We try to do this as scientifically as possible, although, as you’ll see in a bit, there were a few unavoidable flaws…

If you’re already familiar with the tools, click here to jump to the KD Accuracy Experiment.

The rest of you can hang on to your butts…

Part 1: Reviewing the Tools

We’re going to do some quick reviews of the features, pros, and cons of each of these tools, of course, but you can find that information on any blog.

We figured it’d be more useful if we did some actual keyword research on video.

To do that, let’s create an imaginary site in a real market. ​

Today, we’re going to imagine we’re starting a site in the scuba niche.

This is a niche with plenty of opportunities for lots of different kinds of keywords, which gives us a good springboard to launch into a keyword research campaign.

If you want to learn how to find a profitable niche, check this podcast out.

1 – Long Tail Pro Review: Sample Keyword Research Campaign + Pros & Cons

Long Tail Pro

Long Tail Pro is an old school keyword research tool originally developed by Spencer Haws from Spencer just recently sold an 80% stake to Hayden Miyamoto (NoHatDigital) and George Do (Wired Investors).

This is a highly specialized, firmly entrenched, old-school keyword research tool that’s smack-dab in the middle of a new phase of aggressive iterations.

Long Tail Pro allows you to create projects and save keywords. This seems like a fairly simple feature, but it makes all the difference, especially if you’re an independent site builder managing multiple sites or researching multiple markets.

For every seed keyword, you can set both positive (words to include) or negative (words to exclude) filters. This allows you to generate lists of highly specific keywords with laser-focused intent while automatically eliminating keywords you don’t want to see.

Long Tail Pro is the cheapest of the keyword research tools—mostly because you’re not buying a bunch of other stuff with it.

Long Tail Pro gets data directly from the Google keyword tool which ensures the numbers are ‘more’ accurate than the competitors usually when it comes to traffic estimation.

Long Tail Pro works as intended about 70% of the time. The other 30%, it returns errors, freezes or fails to return results. It’s also horrendously slow compared to every other tool. It’s by far the buggiest of all the tools.

Why in the hell does an SEO tool need to show the average number of Amazon reviews? Spoiler alert: it doesn’t.

This feature was crow-barred in because the previous owner was personally involved with Amazon businesses despite its rather obvious uselessness as part of an SEO tool. This feature is better suited for a tool of its own. The market mixing ain’t workin.’

By and large, Long Tail Pro does a good job of analyzing the top 10 results in the SERPs, but there’s no other way to do other kinds of keyword research, like reverse engineering your competitors.

What do I think? Long Tail Pro really does hold a special place in my heart. I cut my teeth on this tool, and I still use it every now and then.

It’s also doing a lot of things the other tools aren’t–namely, creating its own, proprietary algorithms to calculate keyword difficulty and allowing you to manage all your keyword research projects inside the tool, making it a true start-to-finish keyword research software.

That said, it’s got to be careful not to be left behind.

Every other tool here is very fast and on the cloud (supposed coming soon), and the best ones offer ways to conduct deep competitor analysis. There are plenty of ways to do keyword research, and Long Tail Pro currently only allows for one.

Update (June 2016)

In a private conversation with me, George disclosed that Long Tail Pro is in the process of moving to the cloud, and the transition should be complete within the month (he gave me permission to publish this info).

2 – Ahrefs Keyword Explorer Review: Sample Keyword Research Campaign + Pros & Cons

ahrefs logo

In the last year or so, I’ve slowly started to find myself using Ahrefs – as a comprehensive SEO suite–a lot more; currently, it’s my primary tool.

Read our full review of Ahrefs here.

I’ve used it extensively for both link building and competitor analysis, but I’ve not really gotten my hands dirty with the new keyword research functionality, so lets put ‘er through her paces.

Ahrefs is probably the best combination of both types of keyword research: finding and reverse engineering competitors and more “traditional” keyword research (i.e. pulling in a bunch of keywords and data based on a few inputs). They excel at both, and no other tool combines the two like Ahrefs.

When talking to Ahrefs’ Tim Soulo the other day, he mentioned that Ahrefs has many thousands of servers and caches as much data as they can. I’m not surprised. Ahrefs is lightning fast and returns boatloads of data.

This is where Ahrefs really shines, and it’s probably worth buying a subscription purely for its competitor research capabilities. Not only does Ahrefs make it very easy to find competitors, it’s allows you to easily see both your competitors top pages and top keywords.

Ahrefs’ “Content Explorer” is a really, really good way to see what the most popular content in your market really is. I typically don’t use this to find low-competition keywords, but I use it all the time to build linkable assets, and I think of this as closely related to keyword research.

To work effectively as a keyword tool, there’s got to be a way to create a project or a campaign—or, at the very least, to save keywords. Ahrefs does allow you to export keywords, of course, but it’d be great if this kind of functionality was built into the tool.

Ahrefs search volume don’t exactly match with the volumes you will find in the Google keyword tool. Thet seem to be sourcing their data from a 3rd party provider and while overall the numbers seem to make sense, they are not Google’s official numbers.

Ahrefs is not cheap. It’s not the most expensive tool here, but it’s up there. I do think the price is justified, since Ahrefs is really a suite of tools, and they’re all very good. Still, for newer site builders, it can be cost-prohibitive.

What do I think? It seems obvious to me that the Ahrefs team wants this to be the one tool you need for SEO, and keyword research is an important part of their plan for world domination.

And, honestly, they’re scarily close to making it all click. ​

We’ll take a deeper look at their keyword difficulty metric (with all blemishes) below, but the functionality of this tool is stellar.

It’s fast. There’s tons of data. The filters are great. And, perhaps most importantly of all, when combined with the competitor analysis you can conduct with Ahrefs, it basically becomes keyword fire hose. ​

3 – Moz Keyword Tool Review: Sample Keyword Research Campaign + Pros & Cons

moz logo

If you don’t know Moz or use at least one of their tools, you’ve probably been living under a rock.

Their “new” keyword tool is really only newish, since Moz has made what I think is a failed attempt at keyword difficulty in the past; it just wasn’t part of a “real” tool. But they’re officially jumping into the fray here, so let’s see how they do.

Moz’s “Opportunity Score” is a really clever addition to their keyword tool. Essentially, it’s a metric that measures SERP CTR and takes into account all the other weird crap Google’s putting in there these days (answer boxes, carousel ads, etc.). In my view, this is an extremely savvy addition to a keyword tool.

Moz’s “Potential Score” is another new metric that essentially combines all the others. Most importantly, probably, it combines both keyword difficulty and “opportunity.” I haven’t done enough testing to know if the metric is accurate, but that’s not the important thing. The important part is that Moz is making an attempt at helping you understand the ROI of keywords, which none of the other tools in this list do.

Of all the tools here, Moz probably has the best UX. It’s clean. It’s easy to understand. And it’s very simple.

When first looking at suggested keywords, I was shocked at how small the list was. Then I started scrolling and discovered I could scroll more or less indefinitely. The tool looks a bit shallow, but it really does return a lot of data.

This is actually a pretty big oversight. There’s a lot of data, but there’s no way to filter it–at least not in the way anyone would find useful.

You can only filter keywords by volume and semantic relationship. You can’t filter by CPC, other words, difficulty, or anything else.

So, instead of quickly finding what you need, you’re basically force to scroll infinitely and look at each keyword, making the research process incredibly tedious.

By MVP, of course, I mean minimum viable product. This is what I would consider to be very close to the bare minimum of features. There’s no way to do competitor analysis. There’s no way to save projects. There’s no way to filter keywords. And because of that…

Moz is doing about half the stuff the tools are doing, but they’re charging between 500% and 1,000% more. It’s just a teensy bit offensive.

What do I think? There’s a lot to like here. The UX is great. The new “opportunity” and “potential” scores are wonderful additions that show a lot of savvy in the current SEO climate. And the data is deceptively deep.

However, it’s missing some obvious, crucial features.

For example, you can’t spit out thousands of keywords and not really provide a meaningful way to filter them. That, in my view, is a massive oversight. ​There’s also no way to do competitor analysis or ways to manage keyword research campaigns.

Add to that the ridiculous price tag, and it starts to feel silly.

That said, I do feel like this tool has loads of potential energy, and Moz certainly won’t have any shortage of customers, so I’m hoping they can iterate some common-sense features in as well as iterating their price down. ​

4 – SEMRush Keyword Tool Review: Sample Keyword Research Campaign + Pros & Cons

semrush logo

SEMRush has been a go-to competitive analysis tool for a long time, and it’s always done that really, really well… BUT it’s always come in second in most other areas.

How’s their new keyword research tool fare? Let’s give it a whack.

In addition to their very useful, very intuitive competitor map, SEM Rush returns a lot of great data on competitors. While I think Ahrefs has more robust competitor analysis tools, SEM Rush probably provides the easiest way to understand what’s happening in your market.

SEM Rush’s keyword research tool tells you specifically which search features are present for a given keyword. You’ll be able to see at a glance if there are answer boxes, reviews, etc. It’s not as easy to use as Moz’s “opportunity score,” but it’s more detailed. You can also filter by search features, which is a really good idea.

When you start with a general query—say, “sports”—the SEM Rush keyword tool will give you a list of possible categories (e.g. car sports, sports news, sports teams), which can help you narrow down the field significantly.

Similar to a few others in this list, there’s no way to save projects or keywords inside the tool itself, which, again, seems a lot better than managing a bunch of spreadhseets.

It often feels like you’re opening dozens of tabs here. Any time you click on a keyword, you’re taken to a tab with detailed analysis. And, while I appreciate including a detailed analysis, every other tool puts this information in an expandable content toggle below the keyword—it’s just a much cleaner design.

What do I think? This tool is so close to being really good.

In particular, SEM Rush handles generic queries in an ingenius way, auto-categorizing them to allow you to break down your research into sub-campaigns on the fly. That’s a brilliant feature.

They also pull in search features (e.g. answer boxes, carousel ads) for each keyword, which, while not as easy to use as Moz’s similar data, is more detailed, allowing you to get into the nuts of bolts of your potential ROI.

Still, the tool feels very clunky, and, as is the case with Moz and Ahrefs, there’s no way to manage campaigns inside the tool.

Why didn’t we talk about KD score in these reviews?

I’m with you. It’s important.

In fact, in my view, keyword difficulty is one of the most important parts of any keyword research tool (I’ll tell you why below).

However, these new keyword difficulty metrics are complicated, and they’re all calculated differently. But because none of the companies here were willing to share the mechanics of their metrics, it’s tough to discuss them properly.

There’s just not enough data. So, we set out to create our own data by conducting our own experiment. Here are the results…

Part 2: Testing Keyword Difficulty

The Great Keyword Difficulty Experiment: Who’s KD Score Is the Most Accurate?

This might be the most important bit of this whole post.

It’s easy to find a bunch of keyword ideas and their search volumes. I mean, Google’s keyword planner does it for free.

What’s hard is determining how difficult a given keyword is to rank for.

In my view, getting keyword difficulty right is the single most important thing for any keyword research tool.


Because if keyword difficulty is wrong, it can lead to massive amounts of wasted resources.

It’s the difference between spending $5,000 on content (or whatever your budget is) and actually making a profit on it or…. not.

To me, that’s The Ring-level terrifying.

A consistently inaccurate KD score can literally wreck your entire ROI if you rely on a tool to assist you with keyword research.

In other words, the stakes for something as seemingly simple as a KD metric are really, really high, especially if you’re spending a bunch of money to scale content production like we are.

So, obviously, it’s important to see how the KD metrics of each of these tools stack up to each other.

To do that, we wanted to be as scientific as we possibly could, so it’s important to outline our methodology. But first…

The Massive, Glaring Flaw in this Test (…& why it’s not our fault)

Update (June 2016)

As of June 2016, this section is really only half accurate. We now have (more) specific definitions of KD from two of the companies–Ahrefs and Long Tail Pro–and we have a fairly good understanding of Moz’s KD metric, thanks to the guidance of Moz reps popping into the conversation here and on Reddit. That said, there’s still quite a bit of grey area, and there’s certainly no universal standard by which these companies define “keyword difficulty.” Detailed updates below.

Before we dive into the methodology we used to test KD scores for all these tools, I want to tell you why it’s completely and totally flawed…

Of the four tools tested, only ONE defines what their keyword difficulty score actually means.

This is absolutely critical information.

What KD scores signifies an easy keyword? What KD score signifies a competitive keyword?

If you don’t know that information, the only way to even use the KD score is to look at hundreds of keywords and compare them against your own manual assessment of each.

And if you’re going to manually assess each keyword, why even have a KD score?

It’s a Catch-22.

Even worse, by not defining what the metrics mean from a practical standpoint—and by implicitly requiring you to do your own analysis anyway—KD is only useful if you’re already a keyword research expert.

That is profoundly stupid.

And that’s not to say I didn’t try to get answers. I really did.

I asked a Moz rep about it on Reddit. He gave me half an answer (which actually did help me a bit) but ultimately referred me to someone else.

Moz Keyword Difficulty Reddit

And we did tweet Dr. Pete, and he said he would get back to us, but at the end of the day, we got no response.

I was also able to chat with Tim Soulo from Ahrefs on both the Authority Hacker Pro Facebook group and over Skype.

I really do love Tim, but after 30 minutes or so, I still couldn’t get an answer out of him. Here’s probably the most representative snapshot of our conversation.

ahrefs kd question

Tim, brother, it’s not my tool.

It’s up to you to define “easy” for your own metric. Not the user.

I also asked SEM Rush about their rankings in the same email chain I’d been using to talk to them about their tool. What’d I get?

SEMRush Kd question


After all that running around, there’s really only one conclusion left: Ahrefs, Moz & SEM Rush don’t understand their own metrics.

So that’s the flaw.

We’ve got the metrics, but we don’t know what they mean.

Obviously, that makes it tough to test them.

I can only imagine the confusion in the users head right now when they look at these metrics in the tools. But we will try to bring some clarity just below.

And because we don’t know which scores each company thinks are easy/medium/difficult, you’ll have to rely on my interpretation of their scores, which I gleaned by simply looking at hundreds of keywords.

It forces us to be qualitative instead of quantitative.

I’d prefer not to do that, and I’d like this test to be as accurate as possible, so I’m calling you guys out: Ahrefs, Moz, & SEM Rush… please pop into the comments to tell us what your KD means! What’s easy? What’s hard?

We’re all ears.

Update (June 2016)

After we published this post—and (if I succumb to self-importance for a second) possibly even as a reaction to this post—Ahrefs released a post of their own: “How To Use Ahrefs “Keyword Difficulty” Score To Find The Juiciest Keywords To Rank For,” which outlines what their KD metric actually means and how to use it.

Here’s the important bit.

Ahrefs KD score exponential growth

Because Ahrefs bases their KD metric on backlink data only, the difficulty of a keyword is a function of how many links you’d need to build to that keyword to rank for it.

The difficulty increases exponentially, but at the low end of the scale, one KD point is roughly equal to one backlink.

With that in mind, Tim from Ahrefs answers the question, “What KD range can be considered ‘easy’ to rank for?”

What KD range can be considered easy to rank for?

Importantly, though, Tim also noted that the scale changes based on (1) how many backlinks posts on your site tend to generate on average and (2) the KD scores of the keywords you already rank for.

Tim uses his own blog as an example…

bloggerjet # of Referring Domains

…and concludes he can confidently target keywords in the KD 0-30 range.

bloggerjet Organic Keyword

The one company who does define their scores is Long Tail Pro. Here’s their infographic.

Update (June 2016)

The graph below is no longer accurate, but it should be again soon.

Long Tail Pro, who’s always gathered backlink data from Moz’s API, recently switched to the Majestic API because Majestic’s link database tends to be both bigger and more up-to-date (I agree).

Because of that, however, the KD algorithm was thrown a bit off kilter. I reached out to George Do from Long Tail Pro to ask how its affected their KD scores. Here’s what he said.

Long Tail Pro KD algorithm

George told me he hopes to have the KD re-calibrated back to the 30-standard within the month.

Of course, that makes it difficult to use their KD metric for the time being. It also makes it seem a bit unfair to judge the accuracy of their tool based on a temporary mis-calibration that they are working on fixing.

For that reason, in the first update of this post, I’m not going to update the Long Tail Pro KD numbers; instead, I’m going to keep the original scores and update them after they’ve finished recalibrating their algorithm. However, I will make a note at the end that Long Tail Pro is temporarily inaccurate.

Keyword Competitiveness Cheat Sheet
Source: Long Tail Pro

Methodology for Testing Keyword Difficulty Metrics

To measure the accuracy of KD scores, we need to compare it against some sort of control.

Since there’s no objective, universal control available, the best control we have is the subjective analysis of an expert (in this case, me).

So, to some extent, you’re going to have to trust my keyword research expertise for this experiment to be worth anything.

Because that’s the case, it makes sense for me to tell you how I judge the difficulty of a keyword.

For me, keyword difficulty boils down to this question: how easy would it be to beat the weakest site in the top three results?

  • If I can find a site in the first three results I think I can beat with no links and standard “good” content, I call that easy.
  • If I think I can beat it with some links, I call that a medium-difficulty keyword.
  • And if it would take lots of links and epic content, I call it hard.

There are more nuances, of course, but that’s basically how I determine difficulty in a case-by-case, subjective analysis. Here’s what that means for this experiment…

There are two parts to our methodology.

  1. Find keywords we know are very easy to rank for and compare the KD from each tool.
  2. Find Keywords we know are very difficult to rank for and compare the KD from each tool.

Let’s also define those terms in a more scientific way to give ourselves more concrete reference points.

For this test, an easy keyword is a keyword for which a low-authority site is ranking in the top three spots with very few links (<5) to the ranking page.

An easymedium keyword is a keyword for which a low-authority site ranks with a page that has a few good links (5-10).

A medium keyword is a keyword for which a medium-authority site ranks with a page that has plenty of good links (10-40).

A difficult keyword is a keyword for which only pages with lots of links (40+) on high-authority sites rank in the top three spots.

I’m also going to provide a brief analysis for each keyword to justify my subjective difficulty rating and address any nuances I think relevant.

By finding keywords we know are either very easy or very hard, and then comparing the KD of the tool, we can determine the relative accuracy each.

When we do find inaccuracies, we’ll then make our best guess at the reason for the discrepancy.

What if MY analysis is wrong?

That’s the good news.

I really doesn’t matter if my analysis is off or not.

It’s mostly there for (1) convenience and (2) to use as a reference if you don’t have much experience doing keyword research.

If you do have experience doing keyword research–and, in particular, if you feel as if you’re better at it than I am (a strong possibility with this audience)–you can simply do your own analysis of each and compare it to the data we’ve gathered for you without having to buy and test the tools.

What data does each tool use to measure KD?

To better understand our results, let’s take a quick look at how each tool measures KD (quick note here: Long Tail Pro’s KD metric is called KC, or “Keyword Competitiveness,” but it’s the same thing; in this article, we’ll be using KD).

Long Tail Pro

Long Tail Pro created a proprietary algorithm to measure KD. It accounts for several well-known ranking factors: Page Authority (PA), Domain Authority (DA), number of equity (or “juice”) links to the page, the presence of the keyword in the title, and the age of the site. For link metrics, Long Tail Pro uses the Moz database.

Update (June 2016)

Long Tail Pro just switched from using Moz link data to Majestic link data, although the tool still does seem to include PA and DA (Moz) metrics in their data set. Also, as mentioned above, in light of this transition, the ranking algorithm is currently in flux.

Moz KD Score Explanation

Moz’s keyword tool uses PA (exclusively?) to calculate page authority. Moz showed up on Reddit to explain their KD metric further:

Moz KD Score Explanation

Update (June 2016)

As you can see in the comments below, while Moz does use PA exclusively to calculate KD, PA itself is calculated based on lots of metrics–not just link data, making it slightly more robust and flexible than we initially assumed.


Ahrefs KD is based almost entirely on their URL Rank (UR), which is mostly based on page-level links. Here’s more from Ahref’s Tim Soulo responding to our good friend Sebastian in the Authority Hacker Pro Facebook group:

Ahrefs KD

And here’s a snapshot of Ahref’s actual research into the correlation between UR and keyword rankings.

correlation between UR and keyword rankings Ahrefs research

SEM Rush KD Score Explanation

SEM Rush strays from the pack a bit here. Their KD metric is based on the rank of the domains in the SERPs. Here’s the email I sent them and their response.

SEM Rush KD Score Explanation

No comments on that, let’s see how it fares in the test.

A few quick notes before we run our test…

First, I wanted to provide Moz’s own article as a quick reference for you: Search Engine Ranking Factors 2015, which is still the most comprehensive study on the factors that go into a page ranking or not (despite a few imperfections).

Give this a once-over before you read any further. I think it’s important for you, the reader, to know why pages rank or don’t rank, so you can have that stuff top-of-mind as we go through our test.

It’s also a good reminder if you’re going to be doing your own analysis of each keyword instead of trusting mine.

Second, I want to point out that these keywords were picked specifically for their difficulty–nothing else.

In this kind of a (mostly) scientific setting—one in which we are only testing the accuracy of each keyword tool’s KD metric—the profitability, search volume, commercial intent, etc. isn’t necessarily that important.

Actually, one more note…

For three of the four companies in this experiment, KD is a new metric. Ahrefs’ KD score, for example, is still technically in beta testing.

So, please don’t judge them too harshly. ​

I’ll be pointing out some flaws and inaccuracies below, but it’s with the full understanding that we’re working with first iterations in most of these cases.

Instead of a “competition,” I’d like for this experiment to serve as a sounding board for both these companies and you (their customers) to really see what’s happening with KD and how it can be improved.

In the end, the more accurate KD becomes across all platforms, the better it’ll be for everyone.

Now, finally… let’s look at some keywords.

Keyword #1: “bowling tips for seniors”

Perrin’s Difficulty Rank: EXTREMELY EASY​

Quick Analysis: This is probably the easiest keyword I found. The #1 page has no links, and it’s on a low-authority site. The content is very relevant, but at 500 mediocre words, it’s hardly impressive. If I saw this keyword in my research, I’d chalk it up as a shoe-in.

Competitor & Ranking Factor Stats


  • SERP Position: 1
  • Page Authority: 21
  • Domain Authority: 20
  • URL Rating: 10
  • Domain Rating: 39
  • Referring Domains: 1
  • Keyword in Title: partial match
  • Content Length: 536

KD Score: 1

Accuracy: Accurate

Obviously, this is very low. I’m actually a little weary of a ranking this low, however. But it could be my own bias, since most keyword research tools I’ve used don’t go all the way to the bottom of their scale. Still, if there ever was a KD1 keyword, this is it.

bowling tips for seniors KD Ahrefs

KD Score: 23

Accuracy: Mostly accurate

In my years of using Long Tail Pro, I’ve found plenty of keywords that were more difficult than this one with much lower KD scores. That said, if I were using LTP for my research, KD of 23 would still be enough to get me to buy an article, which is why this gets a mostly accurate rating.

bowling tips for seniors KD Long Tail Pro

KD Score: 1

Accuracy: Accurate

Me, Moz, and Ahrefs seem to be on the same page here. This is a super easy keyword. However, it does show you the relative weakness of the “opportunity score,” since this keyword would never make any real money, and the lack of CPC in that particular metric makes some results (like this one) misleading.

bowling tips for seniots KW Moz

KD Score: 63.33%

Accuracy: Moderately accurate

Although SEM Rush’s score is still a bit confusing, this is one the lower scores I’ve seen. It definitely begs the question though: if this keyword can’t even get below 60%, what does SEM Rush think are the truly low-competition keywords.

bowling tips for seniots KD SEMRush

Most accurate for this keyword: Moz & Ahrefs.

All the tools produced low enough scores to get me to buy an article, so, really, they all pass this round. That said, Ahrefs and Moz were the only two that saw the same thing I did: that this is a stupidly easy keyword.

Keyword #2: “nerf N strike elite review”

Perrin’s Difficulty Rank: VERY EASY​

Quick Analysis: This is a more or less perfect example of a really easy keyword. The site in the #1 spot doesn’t have much authority, and the page has exactly zero links. It’s also not very strong content (just over 800 words with very few images and no external links aside from its affiliate links). I would be very, very comfortable competing for this keyword.

Competitor & Ranking Factor Stats


  • SERP Position: 1
  • Page Authority: 18
  • Domain Authority: 28
  • URL Rating: 8
  • Domain Rank: 45
  • Referring Domains: 0
  • Keyword in Title: partial match
  • Content Length: 843 words

Ahrefs is showing a few bugs here, and I’m not sure what no score means. However, by comparing the KD to that of a similar keyword, it seems safe to assume that it means this keyword’s KD may be so low it’s just not on the scale.

nerf N strike elite review KD Ahrefs

In general (and as you’ll see below), Long Tail Pro is often very accurate. Here, though, LTP sh*ts the bed.

This is an absurdly high score for this keyword.

Why is LTP so off? I think it’s mostly because 9/10 sites in the SERPs have crazy metrics. What the tool doesn’t take into account is that none of them are doing what Google wants them to do, so it’s happy to rank a more ideal page.

nerf N strike elite review KD LTP

Moz did a better job with this keyword than I expected, given their strong emphasis on PA (I figured the “big” sites would skew the result more). Still, this is a medium KD for an easy keyword.

nerf N strike elite review KD Moz

These are the kinds of keywords that really show the holes SEM Rush’s KD algorithm. If you base KD only on the power of the domains, you’ll miss the sublties needed for keyword research–and you’ll even miss them by a hell of a lot more than the tools using one page-level metric.

nerf N strike elite review KD SEMRush

Most accurate for this keyword: Ahrefs

Ahrefs is the only tool that reported what I think is an accurate KD here.

More importantly, this keyword illuminates the weak spots in a few of the algorithms–but especially that of SEM Rush, which only uses domain-level metrics to measure the difficulty of even the weird little keywords like this one.

Keyword #3: “best archery range finder”

Perrin’s Difficulty Rank: FAIRLY EASY​

Quick Analysis: This keyword doesn’t appear to be as easy as some of the others, but it’s still very easy, and I’d be comfortable competing against the page in the #2 spot. It’s a low-authority page on a low-authority site.

That said, the content is really strong and extremely relevant, which makes it marginally more difficult to compete against. The site in the #1 spot ( also isn’t unbeatable, since it only has three referring domains and pretty mediocre content, but it’s on more of a medium-authority domain (DR52).

More importantly, I think it’s prudent to compare this keyword to a truly easy one like “bowling tips for seniors,” which, in my view, is one of the easiest I’ve seen.

This keyword is pretty easy, but it’s not the fish-in-a-barrel that “bowling tips for seniors” is, which is something to keep in mind as we analyze the results below.​

Competitor & Ranking Factor Stats


  • SERP Position: 2
  • Page Authority: 16
  • Domain Authority: 20
  • URL Rating: 9
  • Domain Rating: 40
  • Referring Domains: 2
  • Keyword in Title: exact match
  • Content Length: 3,962 words

Although this is a fairly easy keyword, a KD score of 3 seems a bit low. For this keyword, my read is this: it’d be easy to get on the first page but slightly more difficult to break into the top three spots.

best archery range finder KW Ahrefs

If you’re not familiar with Long Tail Pro, a KD score of 15 is very low. So, LTP gets a similar score to Ahrefs here. I think it’s just a smidge too low–mostly because of the on-page factors and the strength of the domain in the #1 spot.

best archery range finder KW Long Tail Pro

Like the others, is a bit too low here, too, I think. Again, this is an easy keyword, and I’d be comfortable competing for it, but it’s not that easy–at least not to break into the top spots. A good way to look at this might be to ask, “is this keyword easier than ‘bowling tips for seniors?'” I don’t think so, and so I don’t think they deserve the same score. best archery range finder KD Moz

This is a difficult result to read because it’s a very low score (for SEM Rush, anyway), and SEM Rush has been relatively inaccurate so far. Still, this is an easy keyword, so a low score is appropriate; however, this is lower than what I think is the easiest keyword, “bowling tips for seniors,” which makes it moderately inaccurate.

best archery range finder KW SEMRush

In my opinion, everyone’s shooting low here. Long Tail Pro seems to be the only tool that didn’t give this keyword (virtually) lowest possible score (which I think is inaccurate), so it wins the round.

Update: Moz’s new score for this keyword and our new understanding of Ahrefs’ KD scale make them the most accurate for this keyword along with Long Tail Pro. ​

Keyword #4: “how to draw Link step by step”

Perrin’s Difficulty Analysis: EASY​-MEDIUM

We’re starting to get a bit harder now, but this keyword is still in the easy range. It’s a very short page (271 words) with only seven referring domains, although the content is hyper-relevant. The site itself is a low-to-medium-authority site.

Here’s the most important thing about this keyword, though: the fact that such a low-authority page is beating a bunch of YouTube videos (not necessarily difficult but are a certain kind of result), indicates that Google really wants to rank pages similar to the one from

In fact, there are a few other sites ​doing so in the SERPs, and even though they’re not as laser-focused, they’re taking up multiple spots.

Competitor & Ranking Factor Stats


  • SERP Position: 1
  • Page Authority: 26
  • Domain Authority: 32
  • URL Rating: 13
  • Domain Rating: 44
  • Referring Domains: 7
  • Keyword in Title: partial match
  • Content Length: 271 words (paginated)

Ahrefs is probably a little low here. This is an easy keyword, but it probably doesn’t deserve the lowest possible score.


how to draw link step by step KD Ahrefs

I’d say this is almost exactly aligned with my assessment, if perhaps a little high. A KD of 30 is very much in the easy-medium range for Long Tail Pro.

how to draw link step by step KD Long Tail Pro

This is also accurate, just a bit on the low side, I think.

how to draw link step by step KD Moz


Not much to say here other than this is not even in the same ballpark as an accurate score.

how to draw link step by step KD SEMRush

Both Moz and Long Tail Pro produced fairly accurate scores for this keyword. Moz was a bit low, I think, while Long Tail Pro was a tad high, but they were still mostly accurate.

Keyword #5: “how to remove wood stain from skin”

Perrin’s Difficulty Analysis: EASY​

Quick Analysis: This keyword is also easy, and the #2 result is, in my opinion, very weak. It’s a relatively low-authority domain (DR48), but the page has only two incoming links, and the “article” is only 118 words long.

Really, this page is more of a forum post (although it’s not technically a forum), which means there are a ton of comments, but most of them aren’t on the page.

This is an example of an article I’d be comfortable competing against with on-page elements alone. We don’t know exactly how semantic search works yet (who does, really?), but I think it’s probably safe to say you can create a more semantically relevant article than this 118-word dink-a-roo.

Competitor & Ranking Factor Stats


  • SERP Position: 2
  • Page Authority: 12
  • Domain Authority: 34
  • URL Rating: 9
  • Domain Rating: 48
  • Referring Domains: 2
  • Keyword in Title: partial match
  • Content Length: 118 words

Same story here as with a few others. Yes, this is an easy keyword, but this basically the lowest possible score.

Plus, while the competitor listed above seems easy to beat, there are some other pages that would traditionally be considered a bit stronger (if not truly strong), like eHow’s page.

So far, one of the more confusing parts about using Ahref’s KD metric is dealing with all the super-low scores.

how to remove wood stain from skin KD Ahrefs

This is a rather high score for a fairly obviously easy keyword.

Aside from the overall KD, take a peek at the KDscores for the results in the 2, 3 and 4 spots. Those pages are all ranking very well with very low KD scores, and I think LTP should adjust for that.

This seems to again illuminate the downside of using raw averages instead of a weight average (​I’m not sure that’s how the algorithm works; I’m just assuming based on these kinds of results).

how to remove wood stain from skin KD Long Tail Pro

I’d call this a very accurate result. This is an easy keyword, but not completely out-of-this-world easy.

If I were to give this keyword my own score based on how I understand Moz’s scale, it’d be very close to the one they generated.

how to remove wood stain from skin KD Moz

Obviously, no result is a bad result. So I’m calling this inaccurate.

how to remove wood stain from skin KD SEMRush


Moz is on the money here.

In my view, these are the most difficult keywords to get right: the easy ones that still might seem difficult.

With these weird keywords, the tools seem to be going either super high or super low. Here, for this example, Moz finds the sweet spot.

Update: Ahrefs seems to be more accurate based on our updated information and data.​

Keyword #6: “horse riding tips”

Perrin’s Difficulty Analysis: EASY-MEDIUM​

Quick Analysis: In my view, this is right on the boarder of easy and medium.

From a pure content perspective, this certainly isn’t the most difficult content to beat. It’s just under 1,000 words, and it’s mostly just generic advice.

However, all of the top 10 results have links. In fact, the competitor listed here (the weakest site in the top three) owns two of those spots, and both of those pages have several incoming links (nine and 26 referring domains, respectively).

So, while it’d be easy to outdo this content, it would likely take a handful of links to break into the top three.

Competitor & Ranking Factor Stats


  • SERP Position: 1
  • Page Authority: 23
  • Domain Authority: 20
  • URL Rating: 15
  • Domain Rating: 41
  • Referring Domains: 9
  • Keyword in Title: partial match
  • Content Length: 989 words

Ahrefs is creeping a bit further up the KD range here, but, in my opinion, it’s a bit low for this keyword.

Most of the more “medium” keywords I’ve measured against Ahref’s KD score have been up in the teens and 20s.

horse riding tips KW Ahrefs

I’d say this is a slam dunk.

In Long Tail Pro, a KD score of 27 is almost exactly what I’d call “easy-medium,” and I image that’s because there are more “medium” domains here normalizing the results.

horse riding tips KD Long Tail Pro

This is close, but I’d still call it too low.

We’ve seen easier keywords with higher scores, and for this particular tool, based on my own analysis, a KD of 7 is very low.

horse riding tips KW Moz

SEM Rush is doing better here than they have so far. Based entirely on previous results I’ve seen, this is a bit low compared to the scores of easier keywords.

horse riding tips KW SEMRush

Long Tail Pro’s KD score of 27 is pretty much exactly what I would have guessed, and it’s right on that easy-medium tightrope, which is where I pegged this keyword in my subjective analysis.

Update: Moz’s updated score improves their accuracy here, and our updated understanding of Ahref’s KD scale adjust their score as well.

Keyword #7: “rock climbing techniques”

Perrin’s Difficulty Analysis: MEDIUM

Quick AnalysisThis is a pretty interesting keyword.

First, I rated this one medium because pretty much everyone in the top 10 in the SERPs has a handful of links, and those in the top three spots are in the double digits (32, 12, and 21 referring domains respectively).

The weakest of the top three, however, is a bit weaker than normal because it’s not content at all; it’s just a tag page, which should be relatively easy to beat. That said, it’s still got links, and it’s still on a strong domain.

It’s also worth mentioning that there’s a ​relatively weaker page in the #5 spot, although it too has a handful of links.

Competitor & Ranking Factor Stats


  • SERP Position: 2
  • Page Authority: 1 (not calculated)
  • Domain Authority: 69
  • URL Rating: 17
  • Domain Rating: 60
  • Referring Domains: 12
  • Keyword in Title: partial match
  • Content Length: N/A

In my view, this is way too low for this keyword.

Remember, to rank for this, you’d almost certainly have to plan on building at least a couple links, and on a 100-point scale (and based on other results we’ve seen) a KD score of 10 probably wouldn’t–or shouldn’t–include results that competitive.​

rock climbing techniques KD Ahrefs

This is another good result for Long Tail Pro. A KD score of 32 is almost exactly medium, which is more or less how I view this keyword.

rock climbing techniques KD Long Tail Pro

I’d say this is low.

In my view, “easy” keywords don’t need many links–this stuff we’re used to going after as independent site builders. With keywords like this one, you’re starting to enter into the link-building territory, and we’re starting to compete with more established sites and pages.

It’s not crazy difficult yet, but it’s certainly not as easy as Moz says here.

rock climbing techniques KD Moz

SEM Rush gets this one mostly right. And, although their KD score is the most difficult to read, 70% seems to be about the median, making this result (for the most part) a winner.

rock climbing techniques KD SEMRush

Long Tail Pro seems to capture the moderate difficulty of this keyword, which is a result of weak content but stronger links and more established domains.

Update: With Ahrefs’ KD scale in mind, their KD score is also accurate here. ​

Keyword #8: “scrabble tips”

Perrin’s Difficulty Analysis: MEDIUM​

Quick Analysis: This is an odd keyword with a weird mix of results.

You might think it tough on your first pass, since you’d be competing with the likes of Hasbro, which is both highly authoritative and extremely topically relevant.

However, there are plenty of low-authority sites further down on the first page, and the #3 result seems extremely weak to me.

Not only is the #3 mostly irrelevant, it’s also a category page instead of an actual piece of content. Plus, it has zero linking domains.

Because it’s so weird that this particular page is beating more relevant pages from sites like, it’s safe to assume there are some other strange forces at play here.

Still, while our litmus test is the “beatability” of the weakest result in the top three results, because of these results are so weird, it might also be worth nothing that we can certainly beat some of the lower results, especially with a few links.

So I’m calling this one medium.

Competitor & Ranking Factor Stats


  • SERP Position: 3
  • Page Authority: 31
  • Domain Authority: 66
  • URL Rating: 11
  • Domain Rating: 61
  • Referring Domains: 0
  • Keyword in Title: partial match
  • Content Length: N/A

As weird as these results are, I still think this is low.

While the URs of a few of the lower pages are probably pulling this score down, to break into the meaningful spots, it’d certainly take more effort that is indicated here.

scrabble tips KD Akrefs

This is pretty medium, according to Long Tail Pro, but based on my own experience with the tool, KD scores above 35 are usually where I start getting a bit uncomfortable.

I’d probably call this score a bit high–based soley on the ease at which I think you could beat the #3 result and the relative weakness of some of the lower pages.

scrabble tips KD Long Tail Pro

This would be an easier call to make if I understood the Moz KD scale better, but my gut tells me this a tad low. Of course, the results are weird, and Moz seems to often land on the low-side of my assessment, so I’m still calling this moderately accurate.

scrabble tips KD Moz

In my opinion, SEM Rush is way off here, likely because their KD is being skewed by the more powerful domains in the top results.

scrabble tips KD SEMRush

None of the tools really lined up with my own analysis, but it’s tough to hold it against them with a SERP this odd. That said, Moz and Long Tail Pro seem to mostly get it right.

Update: Moz’s new score makes them the winner for this keyword. ​

Keyword #9: “teach kid to tie shoes”

Perrin’s Difficulty Analysis: MEDIUM​

Quick Analysis: Honestly, the first two results in this SERP are a toss-up.

The top spot is a YouTube video, and while those can often be easy to beat, Google sometimes shows preference to videos for certain keywords; however, it’s the only video in the results, so that may not be the case.

The strength of the second result depends on how good Google is at determining the semantic and topical relevance of webpages (there’s no exact-match title, but it’s clearly a very relevant page).

The third page is the one I think we could beat with a few links. It’s on a very authoritative domain, but it’s only 299 words long, and it only has a handful of links.

Competitor & Ranking Factor Stats


  • SERP Position: 3
  • Page Authority: 50
  • Domain Authority: 89
  • URL Rating: 17
  • Domain Rating: 71
  • Referring Domains: 8
  • Keyword in Title: partial match
  • Content Length: 299 words

Ahrefs returned an error for the exact keyword here (again, Ahref’s KD is in beta), so I searched similar phrases and used the KD of the closest result I could find.

teach kid to tie shoes KD Ahrefs

This is likely a bit high, and it’s almost certainly because of the strength of the DA scores present here. That said, it’s closer than I thought it would be and still in LTP’s medium ranges.

teach kid to tie shoes KD Long Tail Pro

Moz gets this one right, which is honestly a bit surprising if they’re basing KD entirely on PA (I’m actually starting to believe they’re either adjusting for other metrics or weighting PA somehow).

teach kid to tie shoes KD Moz

SEM Rush is way, way off here. This is obviously not the most difficult keyword we’ve seen.

teach kid to tie shoes KD SEMRush

Based on my assessment, Moz is the most accurate of the four tools here (which honestly surprised me a bit, since they lean so heavily on PA).

Updated: Based on new scores and KD scales, it’s mostly a tie between the three tools listed; however, I think everyone is a tad high.​

Keyword #10: “what to do if your social security card is stolen”

Perrin’s Difficulty Analysis: MEDIUM-DIFFICULT​

Quick Analysis: Here’s where we’d really have to start testing our link-building mettle if we were going to really compete.

At first glance, this keyword may look really hard, but I think it’s more at the upper end of the medium range. Here’s what I’m seeing.

Most of the top results are government sites (scary), but the top result–even though it’s on a very established domain–only has 22 referring domains, which is beatable with a well-run, medium-sized link building campaign.

And I don’t think it’d be massively difficult to build those links, since the content for the ranking page isn’t all that “epic,” despite the enormity of the issues you could cover on this particular topic.

Competitor & Ranking Factor Stats


  • SERP Position: 1
  • Page Authority: 48
  • Domain Authority: 84
  • URL Rating: 22
  • Domain Rating: 66
  • Referring Domains: 22
  • Keyword in Title: partial match
  • Content Length: 817 words

This is almost certainly an error. I checked to make sure other results weren’t malfunctioning, and everything seemed in order.

Still, we can’t use an error, so I’m calling this inaccurate.

what to do if your social security card is stolen  KD Ahrefs

Long Tail Pro is low for this one.

We’ve seen similar scores for much easier keywords–keywords that wouldn’t require a link building campaign to rank.

This keyword almost certainly would, so this KD score is a bit misleading, especially if you’re using their KD scale as a guide.

If I was going to guess, I’d wager the low KD score is likely because the algorithm weights exact-match titles in its algorithm, and it might be weighting them a bit too heavily here.

what to do if your social security card is stolen KD Long Tail Pro

This is a bullseye for Moz.

Based on my assessment of Moz’s KD score, 42 is what I’d call medium or medium-hard, which is in line with my analysis.

what to do if your social security card is stolen  KD Moz

I’d say this is probably high; it’s almost certainly because SEM Rush places such a strong emphasis on perceived domain authority, and these results include a bunch of .gov sites.

what to do if your social security card is stolen  KD SEMRush

Moz seems to be the only tool that puts this keyword in the medium-difficult range, so it wins the round.

Keyword #11: “how to bench press”

Perrin’s Difficulty Analysis: HARD​

Quick Analysis: This is a good example of a difficult keywords that still–perhaps sadistically–gives you hope.

Every site in the top three results–the top 10 for that matter–is an authoritative site with extreme topical relevance.

There does seem to be a teensy chink in the armor with the #2 result, a page with only 28 referring domains and fairly short content. Still, it’s, and the SERPs are leaning toward established fitness sites for this keyword.

Competitor & Ranking Factor Stats


  • SERP Position: 2
  • Page Authority: 46
  • Domain Authority: 82
  • URL Rating: 24
  • Domain Rating: 69
  • Referring Domains: 28
  • Keyword in Title: partial match
  • Content Length: 1,067 words

In my understanding of Ahrefs KD scale, this would be more of a middle-of-the-road score, which is a bit too low for this rather competitive keyword.

how to bench press KD Ahrefs

Long Tail Pro gets this one right as well, I think. I put this keyword just at the bottom of the hard range, and a KD score of 51 says LTP feels the same way.

how to bench press KD Long Tail Pro

This is a tough call. Moz is flirting with accuracy here, but we’ve seen similar scores for easier keywords, and I don’t think a score this low is justified.

That said, because of Moz’s overall consistency, the score would give me a pause and cause me to investigate further, which gives helps it salvage a few points.

how to bench press KD Moz

This is an accurate score from SEM Rush; however, at this point, it’s I’m not sure whether or not they’re more actually accurate with more difficult results, or they just think most keywords are difficult.

how to bench press KD SEMRush

My analysis puts this keyword at the bottom of the “hard” range. Long Tail Pro matched my assessment. SEM Rush was close, but I decided they rate this keyword to be more difficult than it really is.

Update: Based on its new KD scale, Ahrefs’ nudged its way into the top here.​

Keyword #12: “skin care”

Perrin’s Difficulty Analysis: VERY HARD​

Quick Analysis: This is obviously a very difficult keyword.

Aside from the bazillions of sites try to rank for this, every site in the top five has a buttload of links. The site in the #3 spot (the weakest of the top three) has 61 referring domains, while the top two have several hundred between them.

I’d would not at all be comfortable competing for this keyword.​

Competitor & Ranking Factor Stats


  • SERP Position: 3
  • Page Authority: 61
  • Domain Authority: 83
  • URL Rating: 30
  • Domain Rating: 65
  • Referring Domains: 61
  • Keyword in Title: exact match
  • Content Length: paginated list

Ahrefs is spot-on here. This is a very difficult keyword. However, based on this small test, it does seem difficult to reconcile the gulf between what they think is easy and what they think is hard. This keyword is a good example of the high end of that extreme.

skin care KD Ahrefs

I think the last several results show how smoothly Long Tail Pro moves up the difficulty scale. This is not the most difficult keyword, but it’s obviously very competitive and out of the range of a regular Joe. This score demonstrates that.

skin care KD Long Tail Pro

Moz is also very accurate here, essentially getting all the same things right as Ahrefs and Long Tail Pro.

skin care KD Moz

SEM Rush is technically accurate here, but they also seem to be giving every keyword roughly the same score.

skin care KD SEMRush

Everyone agrees here: “skin care” is a highly competitive keyword.

Keyword #13: “lawyer”

Perrin’s Difficulty Analysis: FINAL BOSS​

Quick Analysis: Just for funsies, let’s see how the tools do with the most competitive keyword I could think of.

Competitor & Ranking Factor Stats


  • SERP Position: 3
  • Page Authority: 86
  • Domain Authority: 83
  • URL Rating: 82
  • Domain Rating: 68
  • Referring Domains: 5,538
  • Keyword in Title: exact match
  • Content Length: homepage

Not much to say here; Ahrefs understand this is an incredibly difficult keyword.

lawyer KD Ahrefs

While this is obviously a very high KD, and for most of us, once you get past, say, KD40, the score doesn’t matter much any more, I still would have expected this to be a bit higher.

lawyer KD Long Tail Pro

Not much to say here either; Moz is telling you to run away as fast as you can.

lawyer KD Moz

I don’t even know where to begin, so I won’t start. Remember, this is roughly the same score SEM Rush gave “scrabble tips.”

lawyer KD SEMRush

While everyone but SEM Rush agreed this was a very difficult keyword, Moz and Ahrefs were the ones who really captured the absurdity of trying to compete.

I was actually going to write up results for several more keywords in the “hard” range, but as I ran the tests I saw a trend…

…everyone agrees on the difficult keywords… except, of course, SEM Rush. So I figured it wasn’t really worth writing out a bunch of results if they were all pretty much the same.

The Verdict: Who’s the most accurate?

For my money, the most accurate KD scores belong to Long Tail Pro & Moz. Both were accurate or mostly accurate the majority of the time.

Update (June 2016)

Based on Moz’s updated scores and Ahrefs’ new KD scales, I feel comfortable saying these two tools have the most accurate KD scores–with Ahrefs edging out Moz by a near-negligible margin.

​Long Tail Pro also delivered a strong performance; remember, though, that the data for Long Tail Pro was not updated, since they are currently tweaking their KD algorithm).

Ahrefs Long Tail Pro MOZ SEMRush
# of wins 4 7 8 1

Update (June 2016)

We can see that the new winners are Ahrefs and Moz. However, the really cool thing to note here is just how much the tools have improved across the board. To me, this highlights Moz’s aggressive iteration and both companies’ willingness to respond to user feedback.

Ahrefs Long Tail Pro MOZ SEMRush
# of wins 10 7 9 1

Here is what I thought of the tools while running these tests

Long Tail ProLong Tail Pro would have likely won this little contest outright if it didn’t have such a hard time with easy keywords for which the SERPs were skewed by higher PA values (like “nerf N strike elite review”: a keyword for which a very low-authority page was easily beating a bunch of YouTube videos and Amazon product pages).

Aside from those, however, Long Tail Pro moves up the difficulty scale the most smoothly by far.

Although Long Tail Pro technically had one fewer outright “wins” than Moz, it seemed to more accurate in a lot of cases and closer to the mark more often overall.​

Update (June 2016)

This is no longer accurate. While Long Tail Pro still performs well, it’s currently edged out by both Moz and Ahrefs.


Gael’s Note

One thing I am noting is that Long Tail PRO seems to win more battles in the medium competition range and loses out on obviously easy or hard keywords.

That’s quite important given the fact that most of the “money” keywords you will target will most likely be in the medium competition range.

For me, this gives some brownie points to Long Tail PRO.

I still hate the fact that it’s incredibly unreliable and buggy though.

moz logo

Moz really is a close second, though.

I was honestly surprised by Moz. I didn’t think they would do that well knowing they use PA almost exclusively to calculate KD, since PA relies on Moz’s notoriously incomplete database.

It was so accurate, in fact, that I’ve come to suspect they’re not using PA only. I’d wager they’re either weighting PA scores somehow or secretly using a more detailed algorithm to calculate KD.


Gael’s Note

I usually don’t like MOZ but the fact that they seem so accurate together with the fact that it’s a bug free web app really makes we want to give them a shot for our next content batch.

ahrefs logoAhrefs’ KD score was a bit disappointing, especially since the keyword tool gets very high marks in virtually every other area.

While Ahrefs did come down on the correct side of the 50-yard line most of the time (which means you can use their KD to roughly tell if a keyword is easy or hard), the scores were often extreme or unpredictable.

Update (June 2016)

Based on our tests, Ahrefs is the most accurate keyword research tool with Moz coming in a close second.


Gael’s Note

I love Ahrefs and it’s still my #1 recommended SEO tool but I have to agree with Perrin on the fact that the keyword difficulty while roughly right can be made muuuuuch better.

However, they are the company that iterates by far the fastest so I am confident they will be fixing this soon.

…and then there’s SEM Rush.

semrush logo

If nothing else, SEM Rush’s KD scores should demonstrate how completely misguided it is to judge SERPs on the power of the ranking domains alone.

I love the tool for a lot of reasons, and I still use it often, but, in its current state, their KD metric is utterly broken and you should not be using it for any meaningful decision.


Gael’s Note

For me SEMrush is one of these companies that had a first mover’s advantage which is why so many bloggers and online marketers still recommend / use them.

But their inability to innovate and iterate fast enough made them lose a ton of ground to companies like Ahrefs who clearly overtook them from a technical standpoint.

All this leads me to a burning question…

Update (June 2016)

What follows in the original post is now a bit over-the-top.

The top three companies here all seem to do a good job of calculating KD, and if you already use and like one of these tools, they all perform at par or better.

That said, it should also be clear that none are perfect, and even in our relatively small sample here, we were able to find weird outliers for each tool.

​How do you fix that? In my humble opinion, you fix it by incorporating all known ranking factors based on their correlation to actual SERP results.

There’s no shortage of data. Moz publishes this data every year, and Ahrefs has been running massive experiments based on their own gigantic database.

In other words, thanks to their flexibility and iteration speed, the top tree companies are doing a good job, but it’s still very much a race. ​

Remember: keyword difficulty is a weird, esoteric, relatively new metric that might mean different things for different webmasters​–and the first company to really define an industry standard is going to be the one who really wins.

We really do have a good idea of Google’s ranking factors… so why is only one company using them to calculate KD?

Long Tail Pro is the only keyword research tool who bothered to create an algorithm that actually considered the myriad well-known ranking factors Google uses to sort out the SERPs.

Moz (supposedly) only uses PA. Ahrefs uses UR. SEM Rush only uses their domain rank.


I asked, and these companies have even argued amongst themselves.

Basically, they say stuff like, “In all our tests, UR correlated most strongly to rankings.”

Okay, fine.

Maybe it’s even a great metric.

But is it the only factor? Why not include the others?

I’ve heard similar arguments from all three of the big companies, and frankly, the arrogance is annoying, especially when a few of these KD measurement systems still seem broken.

Part 3: What's the Best Tool?

Before throwing down a judgement, I should reiterate that I really do use most of these tools. In particular, I use Ahrefs, SEM Rush and Long Tail Pro regularly.

As it stands, no one tool is doing everything I need, so I use the best features of each.

I often use SEM Rush to check on competitors and to leverage their competitive positioning maps. I use Long Tail Pro’s KD score. And I use Ahrefs for in-depth competitive research. ​

I’ve also built my own tools using the SEM Rush, Moz and Ahrefs APIs, which are all useful for different stuff.

That said, I do have a favorite… ​

I think the best keyword research tool is Ahrefs. ​

The runner up would probably be Long Tail Pro–mostly because its KD score is so accurate–but it’s seriously hindered by its total inability to perform competitor research, which, in my opinion, is one of the most important parts of keyword research.

Update (June 2016)

Even with all the updates, I stand by this: I think Ahrefs and Long Tail Pro are the best overall keyword research tools. Moz seems to have really boosted its accuracy, but as a tool, it still feels limited.

If nothing else, Ahrefs allows me to do all my keyword research–both competitive and traditional keyword research–in one place, and it’s good at both.

Long Tail Pro still has the best KD metric, but if Ahrefs can get their KD score working properly, it’d be tough to make a case for using any other tool.

And this is awesome, because the Ahrefs keyword tool is very new. Some parts are still in beta. And it’s already incredibly good.

Over to you!

Alright guys. Here we are 10,000 words later.

I know much of what you read here is based entirely on my own subjective analysis, and I know the discussion will get a lot more fun if we can figure out the real mechanics behind the Moz, Ahrefs, and SEM Rush KD scores.

Still, I think we made some good progress, and I’d love for you to hit me with your thoughts on these tools.

Which do you like? Which do you use? What do you think of the new KD scores?

Let me know in the comments!

*Long Tail Pro is updating their KD algorithm on a daily basis. This data has not been updated. We will update this data when they’re finished re-calibrating their algorithm.

Do you want to learn how to build 6 figure authority sites?

Subscribe to join our FREE training and…

  • Learn how to build white hat links to your site without headaches
  • Finally have a proven method to finding profitable niches
  • Get access to our foolproof keyword research methods
  • Learn how to outsource high quality content


  1. What about Hubspot’s Keyword Tool? I use it often because it is the only one I can get KD in bulk from

    1. We haven’t checked them but given the fact that they don’t really maintain an accurate link index, I’m not sure how it could be very accurate. Anyway, we’ll check it out!

  2. Very insightful! You just gave a lot of business intelligence to all these companies :)
    Folks who are new-to-intermediately-experienced with SEO are probably never very sure of their competitor research so I’m glad I read a well rounded review of Ahrefs that highlighted its strengths & contrasted it with the other tools. I’m going to read this mammoth article again I think. Thanks for creating it!

  3. Hey guys. Love this post! I’m a sucker for cold, hard research.

    I’ll def. have to come back and re-read to take it all in, but wanted to give you the thumbs up on it.

    Also, curious why you guys didn’t use in your tests? Personally, TermExplorer is my fave KW tool (and it also has many of the features you dock the tools in this article for). Any reason it was left out?


        1. (though I have not done any real research on how they come up with this score and how accurate it is, only that it usually seems accurate with the niche I am researching.

          It’s out of 10, instead of 100.

          1. Thanks :)

            We were also focusing on the most popular tools. For example, we could have included Market Samurai and SE Cockpit, but at some point, the research gets out of hand.

            Nice to know this one is out there, though!

          2. I have also used Termexplorer in the past and I still have an active account.

            These are the results for the keywords:

            bowling tips for seniors 2.37
            nerf n strike elite review 2.97
            best archery range finder 1.58
            how to draw link step by step 2.07
            how to remove wood stain from skin 1.63
            horse riding tips 3.25
            rock climbing techniques 3.1
            scrabble tips 4.73
            teach kid to tie shoes 1.88
            what to do if your social security card is stolen 2
            how to bench press 5.08
            skin care 9.05
            lawyer 10


  4. Great Article Perrin! I have been frustrated with how slow LTP is and how inaccurate SEMrush seems to be. But haven’t tried Ahrefs because of the cost. I guess I will need to try it now. You mentioned: “I’ve also built my own tools using the SEM Rush, Moz and Ahrefs APIs, which are all useful for different stuff.” What did you mean by this? Did you create spreadsheets to utilize all of these or did you create a piece of software? I would love to know if you have created any google docs templates to help in keyword research (and if you are willnig to share them). PS: I want to tell you, Mark and Gael how much I love the Authority Hacker Pro Community. It has been amazing! Keep up the good work!

    1. I was searching for a comparative review on SemRush & Ahrefs, because lately I’ve been feeling that Ahrefs is outperforming SemRush. I’ve been using SemRush for several years now, and while I love it (especially for competitive analysis, keyword scrapping & paid positions), I’ve been a bit dissapointed with it lately. I prefer Ahrefs for monitoring keywords, positions & movements, and continue to use SemRush for research.
      LongTailPro is also nice, but I never considered it a “primary tool”, just a sweet bonus for more keyword opportunities.

      On a side note: I really appreciate SemRush’s site audit tool, keyword tracking & SEO ideas.

      1. Hey Loana,

        Yep, we used to use SEMRush too but they just rely on tools that are progressively getting outdated and release new ones that are just not on par with the competition. I’d say now Ahrefs is a better place to put your money in.

        I totally agree on LTP too, it’s a good starter tool if you’re on a budget but you quickly outgrow it.

  5. Perrin, I truly want to commend you for speaking the truth about the current state of Long Tail Pro and not pulling any punches. Something that Wired Investors chooses not to allow in their blog comments on their website.

    Thanks for creating this comprehensive post. I’m slightly sad to see LTP is still the most accurate tool out there. Because accuracy is only useful when the tool actually works.

    Market Samurai is only mentioned in passing. Is it considered dead as well? I was just about to try that out again since LTP has been performing so badly ever since v3.0.

    I guess it’s time to stop being a cheap bastard and shell out the monthly money for either a MOZ or AHREFS subscription.

    1. Hey MM, We mostly focused on keyword tools offering keyword difficulty as it’s an essential feature in our opinion. To my knowledge, market samurai doesn’t offer that.

      If you had to pick 1 tool I’d still go for ahrefs just because of the link data, ranking data and the myriad of other useful tool even if their KD needs work.

    2. Hey MM,

      Please don’t be sad about LTP being the most accurate tool out there.

      Because it’s not :)

      Read my comment below where I explain how Ahrefs actually wins this test :)

  6. Nice writeup Perrin & Gael. I was looking forward to seeing this one after your comments with Tim in the group.

    It would be interesting to expand on a particular keyword type, and test a few keywords under that umbrella.

    Eg, test a handful of “skin care” keywords (a keyword they were all accurate for), as well as groups for “rock climbing techniques” and “scrabble tips”, which showed some variations in accuracy.

    Based on your last two posts, it looks like you’re considering a move into the equestrian space. :)

    1. Hey Dan,

      Yep we considered going deeper then we looked at the word count (10k+) and realised most people wouldn’t read that much. You’re free to expand on the tests though ;).

      And nope, no horse riding blog on sight for now.


      1. Headed to Mexico in a few weeks to work with Hayden and the Wired guys, no doubt we’ll have juicy data by the end of summer :D

  7. Great post guys! Kudos for the effort.

    Couple things I wanna add.

    Why was there no mention of search volume? Isn’t this an important part of your keyword research?

    And have any of you tried Keyword Canine?

    It’s dog cheap (excuse the pun) but I’ve found it to be on the button.

  8. I purchased market Samurai when it first came out (at a introductory low price if I remember correctly) and have never felt the need to change to anything else.

  9. Great write-up and super-interesting.

    For newbies like me with just one or two sites making money, LTP is the only real choice. I expect there’ll be a time when paying 99 dollars a month for Moz or whatever doesn’t seem totally ludicrous.

    It is crazy buggy at the moment, but I grew up loading games on a Commodore 64 for 8 minutes and then having to reset and start again when it didn’t take. So if anything the nostalgia is part of the value…

  10. Hello, did you had a chance to check the tool? In their tool called Rank Tracker they have a keyword diffulty metric that I’ve been using.

    Let me know what you think about their data.

    And great article, by the way :)

      1. Of course, you are right about that!

        To be a little more useful, I ran myself the tool and here are the numbers:

        bowling tips for seniors – 24.9
        nerf N strike elite review – 42.4
        best archery range finder – 15.7
        how to draw Link step by step – 43.4
        how to remove wood stain from skin – 33.9
        horse riding tips – 32.0
        rock climbing techniques – 39.5
        scrabble tips – 38.6
        teach kid to tie shoes – 39.2
        what to do if your social security card is stolen – 41.8
        how to bench press – 53.3
        skin care – 58.8
        lawyer – 64.4

        It doesn’t seem too off, I think.

        They explain the factors used here:

        And some of these are:

        Page PageRank/Moz’s PA
        Domain PageRank/Moz’s DA
        Number of sites that link to page
        Number of sites that link to domain
        On-page optimization rate
        Alexa Rank
        Social signals
        Page age

        Just another tool to play with :) Unfortunately I had some problems working with the Long Tail Pro tool (it didn’t want to login to my Google account properly 9 of 10 times), but I will try it again soon.

  11. They can’t explain their own tools! It’s all a big proprietary secret! Their secret formula to get your money lol.

    KD is very difficulty if not impossible to figure out. That’s why I believe these companies should offer way more insight and explanation to the scores they offer. Seems a bit unethical.

    If you can’t stand by and offer details to what the scores actually mean then don’t offer it. Can do more harm than good.

    1. To be frank I just have a feeling that the notion of KWD is new is most of these companies and needs to mature in their own mind for them to give a good explanation of it. Thanks for dropping by!

    2. Hey Bo,

      Please see my comment below! We’ve breaken down Ahrefs KD range based on Perrin’s definitions of Easy/Medium/High and I also showed what each KD score means in terms of (average) number of referring domains.

      Hope Ahrefs KD score is no longer a mystery ;)

  12. Great article guys. For me Long Tail Pro, albeit a little slow sometimes, is really all I need for my keyword research. I love the simplicity of the UI and the KC score.

    Am really hoping that the algorithm is accurate though as I’ve just built 2 sites of niche content around it!!!!

  13. Thanks Perrin and Gael. This is a top-notch article!

    First of all, it is wonderful to see the analysis and then the actual test to back it up with some data. I use LTP myself and it feels good to know it’s somewhat accurate.

    Although, I don’t entirely rely on KD, I also use manual research to gauge whether it’s a good fit or not. However, KD is a great way to get a first glimpse of what to dig into.

    Lastly let me add: For anyone who wants to do a review-style article and is using Thrive themes (their package) THIS is how you use it to its full potential! I just love what you did with all the sections, readability and how you instill trustworthiness.

    I use thrive, and I sure is going to apply these techniques to some of my niche sites. Thanks a lot! keep up the good work!

  14. Great post guys. This is the first time I have been across Authority Hacker. Seems that long content is the way to do it, reminds me of Brian Dean and Matthew Woodward’s methods. Seems that to be competitive in 2016 that posts must be over 2000 words just to start.

  15. Great article. Full of in-depth analysis. I use to use Long tail pro, but I often check again by Moz tool.
    I used to hope to find a tool that can solve all jobs. But now I think again, combining tool together will give the most exact result.

  16. This is so insightful and helpful in helping make a decision on which Keyword Research tool to use out of so many. Knowing that without the proper keyword research, writing content for online marketing is sort of like walking in a land mine field blind-folded. I’ve learn a lot about Keyword research than I’ve ever had a grasp on through AH Pro. This is an excellent article and case study by Perrin. You guys just keep on creating content I believe is controversial to online marketing in a sense. This is has to be the best Keyword research tool review online.

  17. Militsa Chervenkova

    A killing article. LT Pro is by far the one I use for KD.

    I am currently exploring SERPwoo and I am about to write a huge article based on their keyword suggestions and KD in my niche.

  18. Hey Perrin,

    Ahrefs actually beats both LTP and Moz in the accuracy of KD metric. With a HUUUGE advantage :)

    I’ll explain that in a second, but first…

    HUUUUUGE thanks to you (and Gael) for that extensive study of Keyword Difficulty across different tools.

    Ever since we released our Keywords Explorer tool I was encouraging people to do these kinds of tests, so that we could see how KD is used in real examples and give our input on how to read it right.

    So now LET’S GET READY TO RUMBLE!!! :)

    In our Skype conversation I asked you to define “Easy”. Which you did, along with “Medium” and “Hard”.

    I passed your definitions to our team and promised to get back to you with a KD range for each of your definitions.

    And to be honest, that task got buried in the pile…

    But this article made us want to look into it immediately :)))

    So here’s Ahrefs KD range, based on your definitions:

    Easy = KD 0-5
    Medium = KD 6-35
    Hard = KD 36-100

    It’s true that we don’t take into account the power of the domain and only look at the backlinks.

    So roughly, here’s how many referring domains (approximately) you need for each KD score:

    KD 0 = 0 refdomains
    KD 10 = 10 refdomains
    KD 20 = 22 refdomains
    KD 30 = 36
    KD 40 = 56
    KD 50 = 84
    KD 60 = 129
    KD 70 = 202
    KD 80 = 353
    KD 90 = 756

    As you can see, that’s perfectly in line with your definition.

    You said “HARD” means 40+ links, and here’s why we start Hard range of Ahrefs’ KD from 36 – because that roughly means 40 referring domains.

    So based on this Ahrefs KD range I took the time to reassign scores and re-calculate the winner.

    Here’s how I was calculating:

    Accurate: +1 point
    Mostly accurate: +0.7
    Moderately accurate: +0.3
    Moderately inaccurate: -0.3
    Inaccurate: -1

    After adjusting Ahrefs KD range in accordance to your definitions of Easy, Medium and Hard, here’s what I’ve got:

    Ahrefs: 9.8 points
    Moz: 5.7 points
    LongTailPro: 5.4 points

    I should also mention, that I had to remove 2 keywords, where we don’t have KD score and only compared the keywords where we do have it calculated.

    I think that was fair to do, because the purpose of the experiment was to study the accuracy of KD score and not the “coverage”.

    All these “dashes” that you saw next to KD while doing the video meant that we didn’t yet calculate the KD score.

    But we’re on our way!

    We have 4 BILLION keywords in database, please bear with us :)

    PS: it would be fair to mention, that guys from Moz didn’t yet explain their KD range in the way I just explained ours.

    So after they jump in and tell us what’s their KD range for your definitions of Easy, Medium and High – the points should be re-calculated again :)

    1. Have you thought about using a log scale for your KD score so there is more granularity at the low and medium end and less at the tough end where it is less a game of keyword research and more a game of fire power anyway?

  19. I guess some metrics such as KD will always be based on the beliefs of the companies that create their algo’s.

    Plus, it is never an exact science and you should always test yourself as much as possible.

  20. Excellent article, Perrin.

    Thanks for taking the time to review and put your findings together. doesn’t get much of a mention nowadays. I remember when I first started working online, I used this a lot but always doubted the accuracy of keyword volume estimates so proved it with a little experiment using a made up search term and a rank checking tool. Not sure if they’ve improved but wondered if you were aware and, if so, had a view?

  21. Wow, what a fantastic piece of content. Really excellent and informative, thanks for all the effort it must have taken to get it together. I recently signed up to Ahrefs, about 2 months back, but haven’t used it too much yet. Will do now though, glad to hear I’m using the one you recommend.

    I was wondering, in terms of keyword campaigns, which I think was mentioned in one of the videos, are there any resources / articles about that process that you’d recommend?

    I have been doing keyword research for posts a couple of years now, basic to intermediate level, but I never thought of approaching keyword research from the perspective of a campaign. I assume this means organizing your potential keywords in a series of related content posts or articles in a particular area within your niche which you publish with a strategic goal in mind. Sounds interesting … thanks.

  22. First of all, great article. I agree with almost everything. Here are my thoughts on software I’ve used:

    SEMRush – Been using them for 3+ years at my day job (fortune 500 company). Love the new keyword research tool but hate the fact they still cannot analyze on a subdomain level (our ecommerce site is on a subdomain of our corporate site). Ahrefs and others (even Spyfu) have this. I feel like SEMRush will be left in the dust soon.

    Long Tail Pro – I tried to like it but it’s just too slow for my needs. The UX also drives me crazy. I prefer clean/professional/efficient over cute/artsy. Good option for newbies.

    Ahrefs – The best. Love using it although it’s only off and on due to the price. Even the $99/month Lite version seems a bit stripped down compared to the next plan up ($179). Price is what is currently not allowing us to move from SEMRush.

  23. Did you notice that Tim @ Ahrefs just added an article about KD called – “Keyword Difficulty” Explained – looks like it is brand new. I see it inside the Ahrefs system, not sure if he posted it anywhere outside of the system. CR

    1. Yep; Tim’s been in the AH Pro members’ Facebook group breaking a lot of this stuff down with us. We’ll likely be updating the post in a week or two when the dust settles and we have time to do more tests. Gotta love Tim for getting in the mix with us :)

  24. Hello,
    Thanks and this is a great post which is made me look further about all the tools you mentioned. Indeed it made me confuse as the KD results mostly isn’t accurate when I’ve checked those manually on Google’s results.

    Having said that, look like the most accurate keyword tool is you’ve used to find keywords you used on THIS test/case study, whatever it is. What you said as easy, medium, hard and so on is indeed accurate.

    So my question is, what the keyword research tool you used to find keywords for this test? Did I’ve missed something?


  25. You know I’ll always have a soft spot for Long Tail Pro… it was the first IM tool I ever bought, and I used it to build an incredible business.

    LTP is so close to being a winner in the market if they could just get it to run smoothly. It’s accurate, cheap, and has name recognition.

    But all too often, it leaves us cursing at our machines.

    I really hope that George and Hayden can get that turned around in time. Because it looks like that Tim dude from Ahrefs is one competitive fella.

    Your write up mirrors a lot of what I’ve seen in the real-world between LTP and Moz. Thanks for taking the time to put it together.

    1. Agreed Quinton, I also have a soft spot for it. I think a lot of us do who have been in this game for a few years. It was one of the first keyword research tools out there. I built my first site with it; I know Perrin did also. I think they tried to do too much too quick. I hope they get it working; it has massive potential, but basically their entire client base is pissed about its functionality it seems.

  26. First off, holy shit, what an amazingly in-depth post. Well done! I am karmaceutical on reddit so I thought I would respond to your suspicions about Keyword Difficulty…

    > It was so accurate, in fact, that I’ve come to suspect they’re not using PA only

    You are right and you are wrong :-) We do use Domain Authority as well, but since Domain Authority is built completely on Page Authority, it is still all PA! Page Authority takes all of Moz’s metrics which are calculated at a Domain, Subdomain, URL, and Link-Level (by link-level data I mean how much “juice” the link passes) and trains it against real ranking data to build a metric that predicts the likelihood a URL will rank. Thus, it is perfect for building a Keyword Difficulty metric since its whole purpose is to predict ranking. Going from predicting ranking to predicting rankability is a straightforward enough mathematical task.

    > notoriously incomplete index

    This isn’t really that important as long as our index is proportionally relative to Google’s. I wrote a post about this before I joined Moz here ( If you haven’t had a chance to read, take a look. It shows how that small differences in crawl prioritization can lead to hugely different indexes. The deeper you crawl, the more distinct your index. This means smaller indexes (like Moz and SEMRush) are actually more proportionally relative to Google’s. Thus, proportional metrics, like PA, have the potential to be more representative and thus useful for solving problems like Keyword Difficulty. We certainly lose it with things like backlink audits because of our smaller index, but our relative metrics (PA, DA, and the whole suite of MozTrust and MozRank metrics) provide a different kind of value in tools just like KWD.

    > arrogance

    I will end the arrogance here for us. We use PA exclusively because we want to give a Keyword Agnostic keyword difficulty score. Since PA is trained on rankings data, all the other link “factors” are already situated within PA. But, we don’t look at titles, meta, etc. because ours is Keyword Agnostic. This is where other tools might shine against Moz if they were to implement on-page factors.

    > Minimum Viable Product

    I was confused with some of the objections here…

    Filtering: You are correct, we mostly have sorting, not filtering. Before you create a list, you can filter by 6 different styles of related keyword algorithms, filter by volume, then sort by volume and relevancy. Once you have created a list, you can sort by volume, difficulty, opportunity, importance, and potential. You are right that you can’t filter on these features before you add them to a list, but that is because we don’t pre-process difficulty and opportunity scores. We do that on-the-fly for your lists so that you get the latest data. One of the reasons why other keyword difficulty scores might fail is because they are calculated on old SERPs. However, we have heard you loud and clear and we are working on more “filtering” features.

    There’s no way to save projects: Lists are used to save projects. Is this not what you are referring to? I’m literally looking at 6 lists I have saved in KWE right now :-)

    Grouping: We should be launching this soon!

    CPC: We had lots of internal discussions about this but, for the time being, we wont be including CPC data.

    Price: Most of KWE’s users are Moz Pro users, and they were grandfathered in. The price for KWE starts at the equivalent $50/mo although we have yearly licenses. We are looking at this again as it was Moz’s first step into looking at yearly licenses. To be honest, I used to be CTO of an agency and we could make that initial cost on a single keyword research report that would take minutes to generate with Moz Keyword Explorer.

    Anyway, that being said, your write up is AWESOME and I am really impressed with your thorough approach. Thanks again!

  27. I’ve been probing through similar articles to find all-in-one SEO tool. So far it seems that the best way researching keywords is actually combining Google Planner + Serpstat.

    Which brings me to the question:
    Can anyone help me to find good tool for keyword research. Serpstat is good but LITE package is not enough for me, and since STANDARD a bit pricey for me ($89/mo) I’m looking for a tool that has similar abilities like URLs comparison (not domain comparison).

  28. Great article guys,
    I got one question thou. I saw an article from ncihepursuit about when seo rules don’t apply if top 10 websites are all e-commerce. as Google is smart and biased towards store that sell them in tht case. Is it real? Or just a myth?

  29. Hi guys,

    That was a fab piece. Very informative. Well done. Have always enjoyed your coaching calls Perrin at Niche Pursuits and wondered why they’d dried up. Guess I now know.

    I’ve been using LTP for maybe 4 months (bought the full version) and really like it but, as you say, it’s very buggy indeed – to the point of frustration. I love the layout and detail of SEMRush (to which I’ve subscribed monthly for maybe 3 months) but read on another blog that Ahrefs rocks so have just finished the free trial of the Standard plan on that.

    I have always enjoyed analysis and feel the layout and detail of Ahrefs is far more geared towards the more analytically-minded among us. As a one-man-band though I can probably afford 2 services, not three. One reason that I’m not yet giving up SEMRush though is that I quite like the on-page SEO reports it gives. Very detailed (though not always correct IMO) and downloadable in a nice-ish PDF for clients.

    However, I am guilty of analysis paralysis and (though I’ve built one site with LTP which is now ranking) I’m getting really bogged down in my next project!!!

    Anyway Perrin, my questions are, in site #5 you say “this page is more of a forum post (although it’s not technically a forum), which means there are a ton of comments, but most of them aren’t on the page.” I see the comments are paginated. Would you say best practice is NOT to paginated them therefore? Would that give the page more “authority”?

    Also you say you can beat sites #1 and #2. They don’t look THAT bad. I see on-page SEO is poor though. Would your plan be to basically make them more engaging with images, SEO AND more content, or not, because often I think there’s only so much you can say about “How to make a cheese sandwich …” or whatever.

    Really do enjoy your work and would be great if you could please reply.

    Best regards,


  30. LTP has now switched from Moz to Majestic data. KC has changed massively. Nothing from LTP except for a short note in the changelog.

    1. I stand corrected. There’s a post on Niche Pursuits. Because of course that’s the place to put that information after Spencer sold the majority of LTP to Wired Investors.

  31. I use AHrefs and Semrush for keyword research all the time, but I think no one can provide more accurate data than Google itself.
    Their keyword research tool provides the best data as it pulls it directly from the big horse.
    What do you think Gael about this?

    By the way, loved you ultra detailed analysis and providing such a wonderful case study.

  32. If possible can you do another one of these post but with the more unheard of names? I have found one that has a clean UX and kwd analysis, its also priced far less than some of the big names check out kwfinder I’d love to know how accurate their kwd is.


  33. Well, I may be the least knowledgeable of the group here but I use SEO Profiler and it has what seems to be to be a great tool and with my knowledge of the industry I’m in it seems pretty accurate.

  34. Wow, awesome post.
    Took me soo long to read it, but it´s really interesting…
    I was using ahrefs some time ago, but maybe i will try this one again..

    Best Regards
    Haris from Designify

  35. Excellent research on your part, Perrin. Well laid out, complete transparent data!

    Never expected such high inconsistencies with these big SEO brands providing world-renowned tools…

    So, it seems like THIS is the time for a major change, THIS is the time for all the big tool’s pros to sit together, decide a standard and then move forward with their individual businesses. After all, all of these tools are aiming to solve problems of their users. And these inconsistencies are hurting us (as a user) the most…

  36. Great post, thanks!
    Problem with Moz tool is that it’s far from accurate but I’m really looking forward for them to address this as their services always offer good results. SEMrush is my favorite. Using it with SERPstat new tool to have everything covered:)

  37. Hi perrin,
    First of all I really appreciate your effort to make this long experiment on KD as I was really confused with the different stats of different tools.
    You have done a really good job, your stats and study has given many ideas to the big seo companies to make their tools more better for KD`s. Nice to read that Ahrefs win…. I am using it from a month and I am loving it.

  38. I was always convinced that Ahrefs is the best backlink tracking tool. And now their keyword difficulty tool is also most accurate. I am loving the way they are making their service more awesome over time. I am going to get paid subscription soon :)

  39. Oh boy. My brain is spinning now. Still have affinity to Long Tail Pro’s approach though. I think it’s fundamentally correct. As for Ahref, they are already an incredible batch of tools, so it wouldn’t be surprising for them to catch up the keyword difficulty game in the near future.


  40. Hi Perrin

    Wow, best article I found by far on this topic. Well done.

    I am completely new at this KW research and analysis thing and am trying to get my feet wet. A friend of mine gave me his Market Samurai subscription to test it out. He told me this was once the No.1 software on the Market. Is Market Samurai still relevant nowadays or would it mislead me if I had to use it for my KW research. Would you recommend I try one of the other tools and if so, which one would be ideal for a beginner please?


  41. Hi guys,
    Really a awesome post on defining the keyword difficulty. Today I was actually looking for this question, As I was checking through some different tools and all were giving the different KD for the same keywords. I was really confused that moment. Now I’m able to understand how & which metrics they are using and why the KD is different in all tools. Can u give some idea about 2 more tools (KWfinder & keyword revealer) whether I can follow them or not ?

    Thanks & Regard;
    Nurullah Khan

  42. Great article, as I’m still in the process of getting my head around which tools are best for which tasks … and where each task fits in the big picture.

    In your blog post at, you use SEMRush and SECockpit to find keywords to be monetized. Would you still suggest using those for the process you outlined in that post? Or can the same type of process be used with Ahrefs and/or Long Tail Pro?

  43. I found a huge mistakes with SemRush. Their reporting of keyword position of a website is incorrect and needs to be highlighted in the article. It is missing a lot of Data where ahrefs does an excellent job of reporting. I think ahrefs is really great in terms of UX and reporting CORRECT data. Ahrefs reporting of KD is very different and completely different algorithm to produce the CORRECT KD. Majestic SEO I have not used because of UX. I will cancel the other three and keep ahrefs because it is simply the best in my opinion at reporting the data, detail of data, and UX. Yes, I use all three because I have an SEO company.

  44. “After all that running around, there’s really only one conclusion left: Ahrefs, Moz & SEM Rush don’t understand their own metrics.” You should see my face when I read this! Shocking truth.

    Good article!! Good tips for all SEO experts out there!
    I use MOZ KD, and from what I have tested it is the best. I have used some keywords I found out with MOZ and AHREFS (never used long tail Pro’s and I checked SEMrush once, never used it again, lol) and from what I can see on SERPS the Keywords I chose with MOZ ranks better than the Keywords found with AHREFS. Like…much better! I am talking about a Hard and competitive niche- small electronics reviews on an Amazon Affiliate Website.

    Maybe this is because AHREFS was still in tests. But I have my eyes open, I will check their update and try it again.

    I also read the comments, there is something about the Tim comments that make me uncomfortable. I like his passion about his tool, I just don’t see the point in trying so hard to lower the other tools.

    One more thing, when you question about MOZ tool using just their PA, maybe you should think about what is PA, what metrics PA use. Moz says about their PA: “We score Page Authority on a 100-point, logarithmic scale.” and “Page Authority is difficult to influence directly. It is made up of an aggregate of metrics (MozRank, MozTrust, link profile, and more)” So PA is not just a simple metric. (you can read this here:

    Again, great article.

    1. Hey Ana, Thanks for the kind words and I’m glad you enjoyed the article! As for the PA discussion I agree with you it is an advanced metric, however, MOZ is known for being much weaker than Ahrefs when it comes to crawling so they rely on a lot less data than Ahrefs does for their assumption.

      On the other hand though, the fact that Ahrefs essentially only relies on linking root domain builds their own weakness in my opinion so nothing is perfect and it looks like SEO experts are not going to be useless just yet :).

  45. Great Post Perrin and Gael!

    My 2 cents to this amazing conversation thread. I have used Moz and Ahrefs in the past. I love ahrefs because it is one of the best at reporting the data, detail of data, and UX. Alongside Ahrefs, one tool that I think should get a mention over here is – SERPed (

    My reasons for it:

    It divides the keyword research in 4 steps to make sure it gives users the best result:

    1) Ultimate Research – One gets thousands of keywords ideas for the seed keyword. One can filter the result on following parameters – Location, Database, Language
    2) Keyword Analyzer – It checks for the KD and tells users how easy and tough it is to rank for a particular keyword
    3) What Ranks Where – Get the list of top sites ranking for a particular keyword and where do you stand.
    4) Long Tail – Get the long tail keyword Ideas for your selected seed keywords.

    Since it pulls data from Moz, SEMrush and Majestic APIs, one gets the keyword research related data points from these tools as well.

  46. Hi Perrin / Gael,

    I have in mind 2 options:

    1) Buy Long Tail Pro + SEmrush


    2) Buy only aHrefs

    These options are mainly based on their cost (aHrefs lite version costs $99/mo while LTP basic plan ($25) + SemRush ($70) costs almost the same.

    I believe that the combination of LTP + SEMrush will be better than just aHrefs, right? Can you give me an advice here? As I don’t know if both tools are getting accurate results in 2016.


  47. Hi, Gael. Thanks for the insightful resource. Your article mentions that “it’s easy to find a bunch of keyword ideas and their search volumes. I mean, Google’s keyword planner does it for free.” However, Google’s keyword planner has some noteworthy flaws when it comes to reporting volume, including grouping together variations of keywords, such as plurals. Of the tools you’ve reviewed, is there one that you like that a) has accurate keyword volume data on these types of variations and b) provides data on enough keywords and variations to make it worth using? Thanks!

  48. Hi, wow, great skyscraper article…

    Especially your evaluation of the Keyword-Difficulty in SEMRush was sobering since I love their tools.

    So I will rely mostly on the very nice KWFinder for my research… you should really check it out, I would call it a MOZ tool in great design and functionality.



  49. Hi guys,

    Fantastic content. I am currently in the process of deciding on my seo tool set and this post was very helpful. I am building and ranking sites in niches which require fresh data sets for keywords and links. Strictly organic. Its pretty competitive and most sites (and related searches) are under a year old, sometimes only months old. From your comments above, it appears you prefer hrefs as a comprehensive tool. Since I’m very focused on finding just a handful of primary keywords for each site, do you think its also worth investing in Long Tail Pro?

    Also, I see hrefs announced keyword explorer 2.0 back in Nov. Any thoughts on the new features?

    thank you!

  50. I was thinking only if Google tells us its algorithm how it rank keywords, there will be no arguments on which tool is the best! ;-)

    Actually up to now I still couldn’t come to terms how Ahrefs calculates its KD (keyword difficulty). In my opinion it is kind of a little too simplistic. Maybe being primarily a backlinks analyzer, they use no. of backlinks as the main criterion.

    From my experience, SEMRush does not give an accurate picture about its keyword difficulty results.

    I do have a question:

    How are Majestic data compared with Moz data? I mean PA, DA, Trust Flow, etc…

    1. Hah, I doubt Google will ever give us a KD score.

      Majestic data is alright when you get used to it, I’d still put it a notch under ahrefs but it’s probably close second.

  51. Hey Guys – opinions are like bums – we all have one – and theyre all different. To me though I thought you would have covered Rank Tracker by Link assistant – , has more useful features than a lot of the other tools out there (Ive used most of them) and the data is accurate and up to date – based on their own data as well as Google’s index. It’s also NOT cloud based or subscription per month based either – and there is a free version for you to play with.

    1. We recently lost access to the originals but we’re planning on updating them along with this article in the near future, sorry! :(

  52. I just start my blogging career, and I think this article really help me to grow my traffic and increase my website viewer. Thank you for sharing your valuable content. I just waiting for more update about the keyword.

Leave a Comment

Your email address will not be published. Required fields are marked *