Home SEO How To Do Keyword Research in 2017 — Ahrefs’ Guide

How To Do Keyword Research in 2017 — Ahrefs’ Guide

68
0
SHARE

Tim Soulo Tim Soulo Tim is the guy responsible for marketing and product development at Ahrefs. But most importantly he’s the chief evangelist of the company.I’m sure there’s no need to waste time persuading you that keyword research is one of the highest return SEO activities that is essential for your marketing success.

There’s also no shortage of great articles that give you very detailed instructions as to executing a professional keyword research, ranking high for thousands of targeted search terms and vastly improving your traffic from Google.

But here’s an interesting observation: each of these guides will give you a somewhat different set of instructions.

Not that any of them is advising you wrong, it’s just there’s no universal approach to executing keyword research.

It will vary based on:

  • Your website (authority, number of pages, quality of content, etc);
  • Your goals and objectives (branding, exposure, traffic, leads, sales);
  • Your budget, resources and deadlines;
  • Your industry and competitive landscape.

This is why you might find it hard to relate to a random step-by-step guide that you stumble upon.

So I’m going to take a different route and give you a keyword research framework that can be easily adjusted to whatever your goals and resources are.

And I guarantee that the tactics and methods described below will vastly improve your traffic from Google.

At least we’ve grown search traffic to this very blog by 2.4x in one year by doing it:

Seed keywords are the foundation of your keyword research. They define your niche and help you identify your competitors.

If you already have a product or business that you want to promote online, coming up with seed keywords is as easy as describing that product with your own words or brainstorming how other people might search for it.

For example, let’s say you’re launching an online store with GoPro accessories. The Google searches (keywords) that you would first think of are:

  • GoPro accessories;
  • gadgets for GoPro;
  • GoPro add-ons.

That’s a no-brainer, right?

But what if you’re looking to start an affiliate marketing website, and you have no idea which niche to pick or which products to promote?

The challenge of “picking a niche” deserves a big and detailed guide of its own. But generally, there are two ways to approach this:

1. “Monetisation first” approach

Start from exploring available monetisation methods. Pick a product or an offer that you like. And then think of search queries that people might be using to find it in Google.

For example, Amazon has an extremely popular affiliate program. So all you need to do is browse their website until you discover a product (or a category of products) that you’re willing to promote.

Another option is to scout affiliate marketplace sites like ClickBank or CJ, that connect product owners with affiliates.

And finally, just review the products and services that you’re using yourself and see if you can become an affiliate.

2. “Niche down” approach

You can start with a super broad keyword and niche down until you see an interesting opportunity.

For example, I’m going to pick “music” as my super broad niche. Ahrefs’ Keywords Explorer tool gives me almost 5 million keyword ideas for that seed keyword:

In order to “niche down,” I need to focus on longer and more specific keywords that have the word "music" in them. So I will use the "Words" filter to narrow down that huge list of keyword ideas to those with exactly 4 words.

And here’s what I was able to find:

  • “music making software free” – Being an ex-DJ, I know there’s a ton of software for making music. So I could start a review site and cover all the latest releases and updates.
  • “game of thrones music” – People want to download music they hear in movies, TV series, TV shows, etc. And given that new TV content is released regularly, this could be a fun niche.
  • “gifts for music lovers” – I’m sure a lot of famous music bands and leading music labels have a ton of merchandise for their fans to buy. Not to mention musical instrument brands like Gibson, Fender, etc. They must have some affordable gift options too.
  • “music games for kids” – Being a father of an 8-month-old kid, I would totally play some fun music games with him.

These niche ideas are obviously far from perfect, but hey, I spent no more than 5 minutes to find them. Invest a little bit more time and you'll inevitably stumble upon something awesome.

So you have your seed keywords figured out. But that’s only the tip of the keyword research iceberg.

The next step is to generate a mammoth list of relevant keyword ideas, while also getting a good understanding of what people in your niche are searching for in Google.

There are at least four good ways to do it.

1. See what keywords you already rank for

If you own a website that’s been around for a while, you should already be ranking in Google for a few hundred keywords. Knowing what they are is a perfect way to kick-start your keyword research.

A good source of this information is a report called “Search Analytics” in Google Search Console:

These are some of the keywords that my personal blog ranks for in Google

Search Console shows your average position for each of the keywords you rank for and how many impressions and clicks this brings you. However, they don’t show the monthly search volume and you’re limited to 1000 keywords only.

If you need more data, you can try “Organic Keywords” report in Ahrefs’ Site Explorer tool:

The keywords that my personal blog rank for according to Ahrefs

Both Google Search Console and Ahrefs’ Site Explorer have quite a few awesome features to play with, so I went ahead and recorded a quick video of how to use them:

2. See what keywords your competitors are ranking for

The chances are, your competitors have already performed all the tedious keyword research work for you. So you can research the keywords that they rank for and cherry-pick the best ones.

If you don’t know who your competitors are, just put your seed keywords into Google and see who ranks on the front page.

Let's do that with a seed keyword that I discovered earlier, “gifts for music lovers.” I see an interesting website ranking on the front page, uncommongoods.com.

Let's now plug that website in Ahrefs’ Site Explorer and browse the keywords that it ranks for:

Sometimes even a single competitor can supply you with enough keyword ideas to keep your SEO team busy for months. But if you're hungry for more, you can go to “Competing domains” report to find more sites like your competitor.

And we've just closed the "competitive research loop":

  1. Put your seed keyword into Google and see who ranks on top;
  2. Plug their site into Ahrefs to see their best keywords;
  3. Find more relevant websites via the “Competing domains” report;
  4. Go back to either step 1 or 2.

The trick to almost unlimited keyword ideas is to repeat this process over and over.

And don’t neglect tapping into related industries. You might discover a lot of great keywords that don’t necessarily relate to whatever you’re offering but can still bring very targeted visitors to your website.

PRO TIP #1A single page can rank for hundreds (sometimes thousands) of closely related keywords. Therefore, it’s better to focus on the top-performing pages of your competitors, rather than scouting their individual keywords.
Hint: Give the “Top Pages” report a spin in Ahrefs Site Explorer (covered in the video below).PRO TIP #2The chances are, there are quite a few keywords that all of your competitors are targeting, but you aren’t. This is without a doubt your biggest missed opportunity.
Hint: Give the “Content Gap” tool a spin to quickly discover what these keywords are (covered in the video below).

The process of easily researching websites of your competitors deserves a detailed guide of its own (like this one). But I tried to feature some of the coolest tricks in the following video:

3. Use keyword research tools

Good competitor research is often enough to fill your spreadsheet with a ton of relevant keyword ideas.

But if you’re one of the leaders in your niche, that strategy is not quite feasible for you. You have to be looking for some unique keywords that none of your competitors are targeting yet.

And the best way to do it is by using a decent keyword research tool. Luckily, there’s no shortage of them on the market:

Regardless of the tool you choose, there’s no preferred workflow for finding great keyword ideas. Just enter your seed keywords and play with the reports and filters until you stumble upon something cool.

Most tools will pull their keyword suggestions from the following sources:

  • scraping keyword ideas directly from Google Keyword Planner;
  • scraping Google auto-suggest;
  • scraping "similar searches" in Google.

These methods are great, but they can rarely give you more than a couple hundred suggestions. For example, UberSuggest shows only 316 keyword ideas for "content marketing."

There are also advanced keyword research tools (Ahrefs, Moz, SEMrush) that operate a keyword database of their own and therefore will give you vastly more keyword ideas.

For example, Ahrefs Keywords Explorer shows 5,570 keyword ideas for “content marketing.”

You can easily go bananas trying to sift through a keyword list of that size, so we have some great filtering options in place:

  1. Keyword difficulty;
  2. Search volume;
  3. Clicks;
  4. Clicks per search;
  5. Cost per click;
  6. Return rate;
  7. Number of words in a keyword;
  8. Include/Exclude terms.

If you want to learn more about generating keyword ideas via Ahrefs’ Keywords Explorer, check out my full review of this tool.

RecommendationGoogle Keyword Planner is the absolute worst tool that you can use for generating keyword ideas. That’s because it is suited for advertisers, not SEOs. Test GKP against any other keyword research tool and you’ll see how limited their keyword suggestions are.

4. Study your niche well

The aforementioned keyword research strategies are extremely effective and provide an almost unlimited amount of keyword ideas. But at the same time, they kind of keep you “in the box.”

Sometimes, just by studying your niche well (and adding a pinch of common sense), you can discover some great keywords that no one in your niche is targeting yet.

Here’s how to kickstart "out of the box" thinking:

  • Get in the shoes of your potential customers: who they are and what bothers them;
  • Talk with your existing customers, get to know them better, study the language they use;
  • Be an active participant in all your niche communities and social networks.

For example, if you're selling waterproof headphones, here are some of the "out of the box" keywords you might try targeting:

  • how to survive a hard swim practice;
  • how to make swim practice go by faster;
  • what do you think about when swimming;
  • best swimming style for long distance;
  • reduce water resistance swimming.

People searching for these things are not necessarily looking to buy waterproof headphones, but they should be fairly easy to sell to.

We recently wrote a pretty detailed article about this keyword research strategy: 4 Ways to Find Untapped Keyword Ideas With Great Traffic Potential.

And I really enjoyed the advice offered by Dan Petrovic here: Advanced New Tail Keyword Research.

While executing the aforementioned strategies, you'll find yourself sifting through thousands of keyword ideas and trying to decide which of them deserve to be shortlisted.

And to help you separate the wheat from the chaff, there’s a bunch of cool keyword metrics to consider.

1. Search volume

This metric shows you the overall search demand of a given keyword, i.e., how many times people around the world (or in a specific country) put this keyword into Google.

Most of the keyword research tools pull their Search volume numbers from Google AdWords, which was long regarded as a trusted source of this data.

But not anymore. For the past few years, Google has been consistently taking data away from SEOs:

But this time, we were able to get away with a cool workaround – clickstream data.

Thanks to clickstream, we can uncover what Google is hiding from us

By modeling numbers from GKP against clickstream data, we’re able to come up with much more accurate search volumes and un-group keywords that have similar meaning.

Another thing to always keep in mind is the dynamic nature of Search volume.

For example, a keyword like “christmas gifts” will naturally spike around Christmas time while having almost zero search volume during the rest of the year.

To check the search volume trend of a keyword you can use a free tool called Google Trends:

And if you’re using Ahrefs’ Keywords Explorer tool, we have a similar graph built into it:

So the search volume is basically an annual average. And if you’re in doubt about the “seasonality” of a keyword, make sure to check the trends.

But there’s one problem with search volume. It doesn’t always accurately predict the search traffic.

2. Clicks

Let's take a keyword, “donald trump age,” that has a search volume of 246,000 searches per month (according to Google Keyword Planner).

That huge search demand implies that you should get a massive amount of traffic if you rank at the top of Google for that keyword. But let’s see what the search results look like:

A fair share of Google’s real estate is taken by an instant answer to that search query: 70 years.

So does it even make sense to click anything at this point?

Nope:

These “uncommon” search results are known as “SERP features” and there are quite a few different types of them:

  • knowledge cards;
  • featured snippets;
  • top stories;
  • local packs;
  • shopping results;
  • image packs, etc.

Some of them will vastly improve search traffic to your website, but others will steal it away from you.

On the above screenshot from Ahrefs’ Keywords Explorer, you can see that 86% of searches for “donald trump age” don’t result in any clicks on the search results. All because searchers are presented with an instant answer via a Knowledge card.

The Clicks metric is totally invaluable in weeding out the search queries with huge search demand but miserable traffic. And we’re proud to be the only tool on the market to have this metric.

We're also able to show you how many of the clicks get "stolen" by search ads:

On the above screenshot, you can see that a fair share of clicks for "wireless headphones" go to search ads, while clicks for "best wireless headphones" are almost entirely organic.

Sidenote. If you want to learn more about the advanced keyword metrics that we have in Ahrefs, check out this article: Keywords Explorer 2.0 by Ahrefs: Setting new standards for keyword research.

3. Traffic potential

Search volume and Clicks are great metrics to understand the popularity and traffic of a single keyword. But that keyword may have a ton of synonyms and related searches, all of which can be targeted with a single page on your website.

Let me explain what I mean with an example. The keyword “I’m sorry flowers” doesn’t look very promising in terms of search demand or traffic:

The #1 ranking result usually gets no more than 30% of all clicks. This means you can hope for around 60 visits per month if you rank #1 for the keyword “I’m sorry flowers.”

And that is a bit of a discouraging projection, right?

But let’s look at how much search traffic the #1 ranking page for “I’m sorry flowers” keyword actually gets:

On the above screenshot, you can see that it is attracting almost 300 visitors from Google per month. That’s because it ranks for 48 different keywords, and not just “I’m sorry flowers” keyword alone.

We’ve been advocating the importance of long-tail traffic for quite a while (here, here, & here), but it doesn’t hurt to stress it again.

People search for the same things in all sorts of peculiar ways. So a single page on your website has a potential to rank for hundreds (if not thousands) of related keywords.

Here are the keywords that the “I’m sorry flowers” page ranks for, according to Ahrefs:

So it's time to stop evaluating keywords just by their Search volume (or Clicks) alone. You need to look at the top-ranking results and see how much search traffic they get in total.

4. Keyword Difficulty

Unquestionably, the best possible way to gauge the ranking difficulty of a keyword is to manually analyze the search results and use your SEO experience (and gut feeling).

I have covered that whole process, start to finish, in a separate article: How To Gauge Keyword Difficulty And Find The Easiest Keywords To Rank For.

But that is something that you can’t do at scale for thousands of keywords at once. That’s why the keyword difficulty metric is so handy.

Each keyword research tool has their own methods of calculating ranking difficulty score. The one we have at Ahrefs is based on the backlink profiles of the top10 ranking pages for a given keyword. The more quality backlinks they have, the harder it would be for you to outrank them.

To date, there has only been a single study that compares the accuracy of keyword difficulty scores from different tools, and Ahrefs came out the winner in this test:

Recommendation High competition is not always a reason to give up on a keyword. It all comes down to the balance between the business value of that keyword and its ranking difficulty.

Some keywords may be super easy to rank for, but the visitors that they bring to your website will never become customers. So it doesn’t make sense to waste your effort there.

On the other hand, some insanely competitive keywords can be the best thing that could happen to your business if you rank for them. So they’re well worth the investment (and wait).

5. Cost Per Click

This metric is mostly important for advertisers rather than SEOs. However, many SEO professionals treat CPC as an indication of keywords' commercial intent (which actually makes a lot of sense).

One important thing to know about Cost Per Click is that it is much more volatile than Search volume. While search demand for a keyword fluctuates on a monthly basis, its CPC can change pretty much any minute.

Therefore, the CPC values that you see in third-party keyword research tools are nothing but a snapshot of a certain timeframe. If you want to get the actual data, you have to use AdWords.

So you have generated a ton of promising keyword ideas and used the aforementioned metrics to identify the very best ones.

Now it's time to bring some structure to your list.

1. Group by “parent topic”

The days of targeting one keyword with one page are long gone. Now SEO professionals are facing a brand new struggle:

Should I target a bunch of relevant topics with one page or create a separate page for each set of keywords?

We know that one page can rank for hundreds (if not thousands) of relevant keywords. But how much is too much? And how do you know which keywords fit your topic and which don’t?

The way we approach it here at Ahrefs is by looking at the keywords that the top-ranking pages for our target keyword already rank for.

For example, the main keyword of this very article is (obviously) “keyword research.” And I want to know what other relevant keywords I can also rank for along with it.

So I take the #1 ranking page for “keyword research,” put it into Ahrefs’ Site Explorer and sift through the keywords that it ranks for:

And in an instant, I see two decent keywords:

  • keyword analysis – 1,400
  • keyword search – 6,200

This means that I don’t need to create separate pages to target each of these keywords (though maybe they actually deserve it), but try to rank for them with this single post.

How do I optimize my page to make sure that I rank for these additional keywords?

I don’t.

That #1 ranking page doesn’t have even a single mention of these keywords, and it still ranks for them. So if they didn’t bother, why should I?

Recommendation My above advice is quite arguable. I’m sure many SEOs will suggest to actually include all these relevant keywords (and their synonyms) on your page to drive relevancy in the eyes of Google.

Some of them advocate using specialised topic modeling tools (like LSI Graph) for generating semantically related keywords.

We haven’t purposefully tested this method and I have no data for or against it. That is why I recommend you try it yourself and see if you’ll get any measurable results with it.

So that’s the very first step in bringing some stucture to your random list of keywords. You need to find which keywords are semantically and contextually related and group them under a “parent topic” to target with a single page.

Sidenote. Here at Ahrefs, we define “parent topic” as the highest volume keyword that a page ranks for.

2. Group by intent

So you have grouped semantically related keywords by "parent topic" and mapped them to different pages of your website. The next step is to group these "pages" by the so-called "searchers’ intent."

Behind every search query that people put into Google, there’s a certain (and oftentimes very specific) expectation. Your goal is to decipher that expectation upfront, so you could build a page that would perfectly match it.

This could be quite challenging at times. Let’s take a keyword, “roses,” for example. What’s the searchers' intent behind it? Most likely it's one of these two:

  1. See some pictures of roses.
  2. Learn more about this flower.

The best way of deciphering intent behind the search query is to google it and see what comes up first. Google is getting better and better in identifying the intent behind each search query, so the search results usually talk for themselves.

The SERP above serves both these intents with an image strip, followed by a Wikipedia link.

But then you get Guns'N'Roses Twitter profile and a song by The Chainsmokers. What are they doing in the search results for the keyword "roses"?

Well, it looks like Google has identified that it's what people looking for the keyword "roses" want to see.

Once you figure the intent behind your keywords, you might want to map it to the stage of the sales cycle that it represents:

  1. Unaware;
  2. Problem aware;
  3. Solution aware;
  4. Product aware;
  5. Fully aware.

The bullet points above are just one of the many ways different marketers map out the so-called “Buyers' Journey.” Here’s an alternative look at it:

Image taken from customerjourneymarketer.com

Whether you want to map your keywords to any of the existing models or come up with your very own one is entirely up to you. For example, Everett Sizemore from GoInflow.com suggests mapping keywords/topics to user personas. Check out his free template here.

My recommendation would be to stick with whatever makes the most sense for you.

3. Group by business value

This grouping is actually closely related to grouping by intent. But this time, you need to figure out which intent drives the best ROI for your business.

If you’re mainly looking for traffic and brand awareness, you might focus on keywords that will bring tons of visitors but won't necessarily convert into leads or sales.

That’s what HubsSpot does with content on their blog. Take a look at their top-performing articles via Ahrefs:

There are tons of people looking for “how to make a gif,” and HubSpot is generating almost 100k visitors from that single article alone.

But how hard would it be to convert someone looking to create a gif into buying a rather complex marketing software like the one that HubSpot sells?

Very hard.

If you have unlimited marketing budgets, you can fire all cannons at once. But most businesses can’t afford this luxury, so they have to think well about which keywords will drive their business and which ones will only drive their vanity metrics.

Most of the time, marketers will focus on keywords with commercial intent, as these are the ones that drive sales and grow your business. If you don’t know how to identify these keywords, here's a pretty cool guide.

Recommendation Don’t ignore good keywords just because they don’t have commercial value. For example, other than bringing tons of traffic to the HubSpot blog, that article on “how to make a gif” also attracts a ton of backlinks:

Prioritisation is not really the “final step” in your keyword research process, but rather something you do naturally as you move through the aforementioned steps.

While you’re generating keyword ideas, analyzing their metrics, and grouping them, you should be noting the following things:

  1. What is the estimated traffic potential of this keyword (group)?
  2. How tough is the competition? What would it take to rank for it?
  3. How many resources should be invested in building a competitive page and promoting it well?
  4. What’s the ROI of that traffic? Does it only bring brand awareness or actually convert into leads and sales?

You can go as far as adding dedicated columns in your keyword research spreadsheet to give scores to each keyword idea. Then, based on these scores, it should be fairly easy to pick the “low hanging fruit” with the best ROI.

Always remember, it’s not the “easiest to rank for” keywords that you should be looking for. It’s the ones with the best ROI.

Please share your keyword research tips

I tried my best to distill everything I know about keyword research into a single, fairly brief (only 4300 words) guide. And my primary goal with it was to lay out a process that could be universally applicable to any website or industry.

There’s obviously more to keyword research than that. So I would love to pass the mic to you guys & girls and have you share some of your favorite tips and tricks that you didn’t see me mention in this guide.

Let’s learn from each other!

Tim Soulo Tim Soulo Tim is the guy responsible for marketing and product development at Ahrefs. But most importantly he’s the chief evangelist of the company.

LEAVE A REPLY

Please enter your comment!
Please enter your name here