Jeff Baker

Keyword research is either the bane of your digital marketing strategy or the king of it. It’s how you practice keyword research and use that information that separates it from being the former or the latter.

Before we talk about how to do keyword research, we need to talk about how not to do keyword research. And kill a couple bad habits while we’re at it.

What was, no longer is

To understand where we need to be, we have to sprinkle in some context of how things used to be done. Jump into my time machine: We are going back to 2011. Come on, don’t be shy. It’ll be fun!

Here is what Brafton used to do for every client:

The Strategy

  • We created a list of “keyword targets” that were to be used in 200- to 400-word news posts. Our lists were usually 20 to 30 keywords in total, but we really only ever focused on a few of them.  
  • We automatically linked those keywords to product landing pages with our proprietary editorial software. For example, “Companies are increasingly benefitting from content marketing services.”
  • Most humans beings never saw these articles.

The research technique:

  1. We started by building out a brainstorm list in (what was then called) Google Keyword Tool.
  2. We chose keywords that had a good combination of search volume and Google’s competition score. The keyword competition score was only relevant to PPC, but we didn’t know that then.
  3. Once we narrowed the list down to 20 to 30 keywords that had high search volume and acceptable competition levels (which were bogus for organic work – but again, we didn’t know that) we handed them to our editorial teams, and they used them in content.

Here’s a snapshot of my first keyword research project for a client:

Bonus points for seasonality


The results: It worked. And it worked well. The years of 2011 and 2012 marked the golden age of content marketing. We had Google’s number. Below is a snapshot of a standard-ish looking slide we would include in ROI reports with clients that shows keyword ranking improvements using solely the technique outlined above:

Have you thrown up on your shoes yet? I threw up twice writing that out. I suppose this is a confessional of sorts, and I’m subconsciously seeking atonement for my sins. But try to keep perspective in mind; this is how it was done. And for the most part, it worked. Unfortunately, a lot of people still think it works.

Also try to keep in mind that our world of digital marketing is moving at 100 miles per hour. This “how-to” guide will be obsolete in a year. 

Read this, use it for a year, then discard it when it starts hurting you – because it will eventually.

How to do keyword research today

Let’s get started. (You can download the PDF here if you want a doggie bag for this post).

1. Brainstorm

This is where you get to go bonkers with any and all query possibilities. Google’s Keyword Planner Tool does a great job of helping you brainstorm, so open it up in one tab along with your website and a few competitors in other tabs.

Start a list of everything that jumps out at you as a potential keyword, beginning with the title tags for the homepage and core landing pages on your site and your competitors’ sites.

I’ll do one along with you. I’ll pretend Brafton is solely a graphic design company that creates infographics for clients.

Tip: Jot down everything YOU would search if you were in the shoes of your target audience. Step outside your bubble.

By default, Google will pull the exact match keywords you entered into the “Keyword Ideas” tab. Make sure you click “Add to plan” for each of the ideas you entered. This will be the start of your list.

Tip: Ignore the “Avg. monthly searches” and “Competition” metrics. The search volume is inaccurate and the competition is only relevant if you are doing PPC. (Don’t worry. We will revisit these later in another tool.)  

Take a look in the Ad Group Ideas tab. Google will bucket suggestions into a bunch of different topics that you can look into and add to your list. Make sure you do that! The only criteria you should be evaluating with this tool is relevance. Click into each relevant category and add the most relevant keywords to your list.

How is your list looking? It should be huge by now, which is good because we will whittle it down later. Export it to CSV and say goodbye to Google. We are done with them.

Progress: 35% Complete

2. Gather Search Volume

There is no point in writing about something if nobody cares to search it. As mentioned previously, Google is going to be very little help in this regard. If you have access to Searchmetrics, Moz or SEMrush, you will get a pretty good idea of the monthly interest. Upload your brainstormed list into your tool of choice. Note that they will all have different figures, but remember, these are all only estimates.

For this exercise, I will be using Moz.

Beautiful. Now we have a search volume number for our list.

Progress: 50% Complete

3. Evaluate Difficulty

I’m going to argue that this is the most overlooked and crucial step that most SEOs fail to recognize, for reasons I can’t quite understand. When evaluating your keyword list, you need to know whether or not you stand a realistic chance of ranking for that keyword. Without that information, you are shooting blind.

For this portion, you need to wave goodbye to your trusty friend Searchmetrics, because their difficulty scores won’t help you.

Continuing with Moz (because I don’t like importing/exporting giant keyword lists all day), we see that each keyword has an associated difficulty score.

So what is this telling you? This is a score from one to 100 based on the Domain Authority and Page Authority of the first page of organic results for the keyword in question. In English, this number tells you how strong your potential competitors are for the keyword you want to target.

If the Difficulty is very high compared to your Domain Authority and Page Authority, that means you will have to outrank some major powerhouses to rank anywhere meaningful. Example:

  • Using Open Site Explorer, we see that our Domain Authority is 56.
  • The keyword “design infographics” has a Difficulty score of 77%.

This tells us that the average domain and ranking page on the first page of Google results is 48 percent stronger than our domain. Let’s see if that passes the eyeball test:

Yikes. We aren’t ranking on the first page for this keyword, ever. You can’t argue with the numbers: The competition is stronger than us. But on the flipside, the keyword “graphic design service” has a difficulty of 26, telling us that the average result on the first page belongs to a site that is 54 percent weaker than ours.

That’s a little more like it. Oh and look, the term also has decent search volume. It’s much less than some of the other choices, but I would much rather rank in the top five for a term that gets between 501 and 850 searches than rank on the third page for a term that has 5,000 times more searches per month.

Narrow down your list to keywords that have an acceptable search volume and a difficulty score that you can realistically rank for. Ideally, you won’t be selecting anything with an average difficulty score that is higher than your Domain Authority.

Progress: 70% Complete

4. SERP Crowding

Let’s take stock: So far you have narrowed your list down to a set of keywords with an acceptable search volume that you can realistically rank on the first page for.

Are we done? No.

What happens if you rank in the third position for a keyword, but the page is extremely crowded with PPC, news, featured snippets and images? Your organic result gets ignored, that’s what. As a searcher, you know exactly what this looks like. If you don’t, Google “car insurance” and you will get an idea of what a crowded SERP looks like.

Fortunately, the Keyword Opportunity tool gives you a crowding score of one to 100, with one being the most crowded. With this in mind, consider the following two keywords:

Note that “graphic design services” has more search volume, a lower barrier to entry on the first page, and less SERP crowding. This is a better keyword selection.

As you are slimming down your list, consider adding another column for “relevance.” Here you could assign a subjective one  to 10 score based on how relevant the keyword is to your brand.

Progress: 75% Complete

5. Designation

We are 75 percent done. Make sure you stick this thing out to the end because the next step is crucial, and oft-overlooked. As referenced before, “targeting” keywords in your content is a waste of time. Just “using them” throughout your website for the purpose of keyword density is a waste of your time as well, and will result in a neutral-to-negative outcome.

Once you have selected a few keywords, you need to designate exactly how you are going to use them.

We have a problem here: I found a great keyword to target for a core product page, but there is another great keyword I know we could rank for. I don’t want to waste this effort!

No problem! Turn it into a blog post or resource page! That’s the beauty of doing this keyword research; the byproduct of the massive list you started with is a number of great adjunct keywords you can target with other types of content.

Progress: 85% Complete

6. Create/Optimize

Now that you have your small list of keywords, determine whether you need to create a new page or optimize an existing page for each keyword.

Crucial note: When optimizing a page for a keyword, you may be tempted to fall back into the old trap of using the exact keyword too much on the page.

This is what you need to do when trying to rank for a keyword:

  • Consider the intent of a searcher that is searching your keyword. Are they trying to learn something? Are they shopping products/vendors? Are they top-of-funnel or bottom-of-funnel?
  • Make sure your content satisfies the intent of the query. For example, someone searching “graphic design marketing” is trying to learn something, while someone searching “graphic design services” is shopping for a vendor.
  • If your content fulfills the intent of the searcher, you may rank for the keyword you are targeting, and for peripheral keywords.

Google intends to return results that will satisfy your intent rather than results with the keyword you searched.

You may be asking, “Well why did we do all this keyword research if the keyword doesn’t matter?

Good question, and the answer is that it would be more accurate to say that we did “Search Intention Research” rather than “Keyword Research.” But don’t be dissuaded, because we had to do the keyword research to uncover the intention.

Now, if we create pages that solve for the intention of the keyword (rather than stuffing it in the content), we will rank for the keyword we want and probably some others. I know it sounds like we are doing this backwards, and we technically might be, but the resulting content will provide the value your searchers are looking for, and Google will reward you for it.

Rather than going off on a tangent about semantic search, I’ll direct you to Neil Patel, who does a great job detailing the topic.

Once you have written your content, use on-page best practices to ensure you have covered the basics. Searchmetrics and Moz both have excellent tools to help with this.

Progress: 95% Complete

7. Refine

Refine. Refine. Refine. Evaluate your new/optimized page in your search visibility reports.


  • The queries your page ranks for.
  • Average position for keywords.
  • Impressions.
  • Click-through-rates.

If the page is underperforming, you may have to try another target, or re-check your on-page optimization.

If the page is performing, now you have a whole new project to start: evaluating whether or not the page is converting visitors and contributing commercial value (but let’s save that for another day).

A lot of work, right? Keyword research definitely is, but it’s what you need to do to be successful, and exactly why it needs to be taken seriously.

Progress: 100% Complete