The foundation of the keyword research component of SEO campaigns always starts out with “gather your keywords, search volume, and PPC costs for those keywords.” But there’s a more robust process that you can build into your organization to get outsized returns from the product of this process.

I say organization, because this should be true if you’re a one-person consulting SEO shop, a small to medium-size agency, an in-house marketer, or just the only person at your business who has been handed “the website stuff.”

Jeff Coyle of Marketmuse, Paul Shapiro of Shopify, Jonas Sickler of Terakeet, and I hash out the “What now?” part of keyword research — what you should do NEXT after you build your FIRST keyword list.

Jeremy Rivera  

I’m the director of special projects at Copypress, working on product management and SEO. We’ve got some awesome content and competitor analysis tools that we’re working on that I’m excited to develop. I’ve been in SEO for 15 years, working with realtors, plumbers, and the whole lot, so I’m happy to contribute and get these guys together to talk about “what to do once you have a list of keywords.”

Paul Shapiro  

Hi, everyone, I’m Paul Shapiro. I lead technical SEO and product management SEO over at Shopify. I’ve been in SEO for a little over 10 years and focus mostly on large enterprise-type clients and in-house roles. Happy to be here.

Jonas Sickler  

I recently moved from an SEO-specific role to a digital marketing analyst position. I’m still at Terakeet. We work with Fortune 500 companies. We recently launched a new category of marketing called owned asset optimization, which kind of combines different aspects of SEO and content and online reputation management, all about prioritizing the most controllable marketing assets that brands have in order to connect with customers.

Jeff Coyle  

Hi, I’m Jeff Coyle. I’m the cofounder, Chief Strategy Officer, and AI content marketing speaker for Market Muse. I’ve been fortunately or unfortunately in the search, lead gen, and product management space for about 25 years, and I’ve been building content strategies for mid-market to small enterprise companies of all types, shapes, and sizes with market views for the last eight years. And looking forward to this discussion.

How do you determine what’s actually going to give you the best ROI?

Sabrina Hipps  

Just to kind of set the scene for you all and for our future viewers, we are going from the perspective of your small company or mid-sized company, or just one person who has gathered this keyword list. And you’re looking at all of these words and it’s kind of like a “now what?” What do you do with this?

How do you dive into it? And how do you determine what’s actually going to give you the best ROI?

My first question is: you’ve got that first keyword list, what would you say is the best first step — just what do you do first after you have that list?

Jeremy Rivera  

My first instinct is to say go back and get MORE keywords. But really, what you need to do is start applying some logic and separating out because you know whether it’s SEMrush, Ahrefs, or any of the keyword research tools out there. What you get out is like throwing an ad into the ocean and you pull it back, you’re going to have a really weird catch. 

You need to have some sort of process in place to start narrowing that down and grouping it in that. Also, you know, talks. You need to have a better sense of, like, how big are you setting your net in the first place? Are you trying to catch all of the keywords for your entire site? Are you trying to scope out your plan for the next quarter or the next year? 

So, setting some guidelines and boundaries and figuring out, you know, how do I group this? What are the words that make sense together? What are the clusters? And I know Jeff loves clusters, and I want to see what Paul and Jonas do in this phase of grouping, labeling, grabbing things, and clustering things.

Jeff Coyle  

I think one thing I’ll add there is that I think it matters how you got to that list, what you do next. So, if you opt it into your favorite solution like the ones that Jeremy mentioned, if you’re starting from a word, turning it into a list, it’s going to be looking at related topics.

Typically, I’m going to be looking for string matches, and then you’re at kind of the universe but the process of universe development may come from your site. So, have you read all the pages on your site that you already have? You [look at] all the things you wrote about plus all the things you’re ranking for. That is the universe of what you want to be or what you are about. And if that’s where you’re starting, maybe you add some things to it.

With the keyword research solution with other seeds, maybe there are things you haven’t approached yet. And maybe you have a weaker side and you’re doing it from a competitive cohort analysis. So, you’re looking at a competitor and understanding what they’re doing. But in the end, you have that universe that you then need to label and start to put into buckets. Then it gets into the “how do you put it in buckets,” but those are some of the ways to do it.

And you know, it’s not just about you. You are a small company, but you know, I think you want to get buy-in on the final product. So, get some inputs from as many people in the organization as you can, so that, you know, they won’t look at that and go, “Well, we don’t want to write about that.”

How do you group and cluster your keywords?

Sabrina Hipps  

Jeff, I’m glad you brought up the expert aspect, as that is something we’ve discussed. Jeremy and I have been fortunate enough to do webinars on this topic before. Sourcing outside experts and information to even just build that initial keyword list and keyword universe is so important.

Making sure you’re really having a robust and varied keyword universe to start with before you start clustering. The next question is okay, so you have a good idea that you’ve built a good sized “keyword universe.” How do you cluster this? What do you start looking at? Do you just start throwing words together in one chunk and go “this looks good,” or is there a way to go about it more methodically?

Paul Shapiro  

Yeah, I mean, it very much depends on what you have going on. It also might depend on like, what is the tenor of your website? Like how long have you been doing this? What about your content marketing efforts? Are those really good? Are we talking about your competitor versus other people being the competitors?

Because if you have major keyword gaps in your content, then maybe the competition is, you know, step one, but then step two is outside of the competition — what is the competition not doing that you’re not doing?

So you need to determine that and then you worry about, are you clustering your content? Then you can layer in metrics like search volume, competition, right to win, and those sorts of things that you start to zero in on, then what are your next steps for content creation?

Jeremy Rivera  

It almost feels like, you know, you have the first step of “I’ve got a list.” And then you ask, “What did I miss?” What wasn’t there? So that’s going back to talking to customer support representatives and talking to actual customers to figure out what their questions were, and then start to fill in from those sides, the human detail. But then also look at your market picture like, am I David or am I Goliath?

Am I the 800-pound gorilla? Or do I have to FIGHT with a 100-pound gorilla (and he’s really jealous about these particular bananas)? I have no chance for THOSE bananas but maybe these, and then we get to finally having some sort of list. Is there anything else? Did I get enough keywords to move onto the filtering process?

Jonas Sickler  

Yeah, to build on what Jeff had said earlier, the idea of how you get here is important because you start with this universe of keywords, but it all comes down to solving the problems of your audience and connecting with your audience.

So every keyword in terms of how to group them always comes back to the idea of when someone searches for this what was the intent behind their search? So the two predominant means of grouping tend to fall into keyword matching, you know, this word contains X, therefore, group it with these things that have the same set of search results in Google. Therefore, they’re solving the same intent and that’s the way I prefer to do it because there might be a lot of different ways to phrase something that is trying to reach the same conclusion.

And so, whenever we’re grouping keywords, I think the important thing is to keep in mind who the audience is, what their journey looks like from the very first discovery of a problem that they’re realizing they might need to try to solve — all the way through to comparing different types of solutions and brands within the solution vertical. Then it’s verifying the trust value at the very end, like, you know, does this brand have good reviews?

So that whole journey needs to be represented throughout the keyword process, and then we would group within that based on similar intents. And whatever methodology you want, I think it needs to be that all the keywords that get grouped into a blog post should be satisfying the same intent of that blog post, no matter if they have different phrasing or similar phrasings.

Jeff Coyle  

Now, just say, I think Jonas hits on it the most, and I think something he said is, it may be different ways of clustering words. So, he mentions clustering based on a data point of — is there overlap amongst search results? So there’s two methodologies that are prominent there.

One is called centroid and the other is called Gomeroi rate. And don’t worry about what that means. It means that you’re going to cluster the words that have similar search results and group them by the one that has the highest volume in the one version. Then you’re going to actually compare all the words against each other.

Imagine a big wrestling match, where everybody gets to like each other. And those are going to allow for the groupings by like to layer in semantic relatedness. So, intent analysis. A good way of doing intent analysis is to look at all of the keyword variants to a word, and they’ll start to tell a story about if there’s an early-stage awareness need, if there’s prices, then there’s the middle of the funnel. If there’s troubleshooting, like queries, you’re going to need that.

You’re going to need local intents if there’s Microsoft, you’re going to need brand intent. It is not saying go write about Microsoft, right? So, you start to learn the buyer journey, the customer journey, and the information journey. Against the story, one would tell if they truly knew everything there was to know about the concept when you get there. The clusters right themselves, and I like to bonus this and say, you start knowing your differentiated value.

Because once you focus on, I know this or somebody in the company knows this, those are the pieces that really keep you differentiated at the end. Start with that so that you kind of have that brick in each of these clusters. It’s the “my team special brick” right as you go and that ends up being a good methodology.

Jeremy Rivera  

And what I kind of hear, too, is you know, I’ve heard it described as like top-of-funnel, middle-funnel and bottom-of-funnel. So, if I was looking at my keyword grid and my list of keywords, I’m almost thinking about adding columns to describe or capture these additional items so that I can kind of bring awareness on a keyword level. Okay, this isn’t just in this particular cluster, but you know, maybe this is a set of keywords that is only for people looking, they’re ignorant of the solution and the problem. They’re up here, versus people who are aware of the problem but don’t know the solution.

And then at the bottom, you have people who know what the problem is, know what a solution is, and are deciding which way to solve it. So, it sounds like we get our keyword list. We asked ourselves, did we make it big enough? How big does it need to be versus our competitors? We start to cluster but we also need to respect some secondary factors about funnels and intent behind the scenes.

Sabrina Hipps  

That’s great. Jeremy and everyone’s insights were awesome. It sounds to me that having a really intimate knowledge of the audience base of your product of what you’re doing is key to this. But in a lot of cases, we may be reaching people who are agency owners who are working with companies that they’re not necessarily ingrained in.

Does anyone have any recommendations on how to go about clustering and filtering and grouping when you might not have intimate knowledge of the actual funnel? Who should you bring in, or should you bring anyone in? Do you have any ideas? 

Paul Shapiro  

I mean, as a recovering agency guy, I would say that’s something you need to do. You need to get that information. You need to talk to people, you need to talk to the in-house marketers, if there are any, or salespeople, or really spend time reading and understanding the product. That’s just part of the game. You have to do that. If you start creating lists of keywords, then say, create content around those, you’re going to look like an idiot. And I’ve seen that plenty of times.

Jeremy Rivera  

Yeah. Yeah, there’s definitely been moments at previous agencies where we came up with a list and we sent it to the client, then they literally laughed, because that’s not even their product. Like, that’s the similar word for different products. Or, you know, there’s so many certain niches that have legal restrictions, AI, or brand restrictions.

That you need to layer on as well like, “Oh, we’re a luxury print brand. So, you know, we’re not a space heater. We don’t want to say swamp cooler, you know, we’re not. We we do air conditioning, but ours is a portable evaporative cooler. We’re a space cooler.”

So, you have to have a feedback loop and mechanism with your vested people. If you have somebody else who’s, you know, signing your checks for this, you need to loop them in and get that feedback. What was your point, Jeff?

Jeff Coyle  

I was just saying it’s important to also differentiate the expertise and the folks that you’re trying to extract the good goo out of their brain, because you have to, if you’re in this role, your job is to be able to ask great questions. Pull as much information as you can, because you’re acquiring knowledge. So, your options are to be a great researcher, right? You find textbooks, you find, you know, business books. So, it is to ask great questions, but you got to when you’re asking great questions, ask do we have to know what they actually have done? Or are they an industry expert? Right?

So, there’s a big difference between someone who understands HRMS software, HR management software, and the market versus somebody who’s actually run the payroll.

Most of those people who understand the software, that software market has never actually run a payroll themselves. You may need both of those people to be able to write a truly valuable, entire definition of a buyer journey.

Your rubric development, and your questions that you ask early on, will make it so that you truly can exhibit expertise and authenticity, even if you aren’t the one who is authentic, which is critical, right? Especially critical right now.

Jonas Sickler  

Yeah, I would say I think that there’s a big challenge and even the question of who you talk to or don’t talk to might be just as important because there’s a tendency to differentiate by saying we’re different. And like you were saying, Jeremy, I’ve seen the same thing where it’s like, “oh, we’re not a cruise company. We’re a voyage company.” But nobody’s searching for voyages unless they’re going to maybe Mars or something.

That’s not happening yet. So, it’s important to figure out not only what the differentiators are of the company, but what language do your customers use when they’re trying to understand the product?

It’s not, what language do you want them to use? Because they’re not there yet. You know, maybe they get there at some point. But you know, it’s really thinking about all the problems they’re trying to solve in a way that somebody who doesn’t understand the product yet is going to use.

So, what would you say, are product people better at than marketers? Or would you say that they’re worse than marketers?

Jeremy Rivera  

They can. That’s a hard one, I think. I think there’s different expertise there. Sometimes it seems like people in charge of product kind of have their own internal jargon and they expect to see that jargon replicated into the marketing but don’t take a step.

That necessary step back to is, like, is this how people actually would pull out their phone and search for this thing? If they speak to their phone and ask for this product? Like, what would it be? What would they actually say?

Jeff Coyle  

I think when you speak with marketers, they often end their journey after the customer buys something because they get paid on leads or sales oftentimes, and product persons will often include the wholesale buyer journey post purchase, post-purchase implementation, onboarding, troubleshooting — whether it’s a pair of socks and you’re buying your next pair of socks, or you know, it’s software or a more complex sale. So, I think you can get something for an industry, different industries, and different stages of the buyer journey from different people.

And so, you know, your sealed C-suite may be very thoughtful and understand the pain points of your sales team, your customer success team, and thinking about post-purchase your marketing team maybe. So be thinking about putting the right person in the right seat for those journeys.

When I’ve seen that be magical is when the sales team actually feels like they were heard when the product team actually feels like their stuff made it to a page because they’re not great writers. Right? And that ends up being the punch line that you are supposed to be.

Sabrina Hipps  

That makes a lot of sense and I think leads perfectly into the next point: what people are searching for? And how it’s impacted by SEO is a lot different than what the product people, the salespeople, might be saying. So how do you really find the value in these clusters? So, you’ve clustered, you cleaned, and you kind of have a list that’s workable now.

But how do you say this keyword of “cruise” is going to be more valuable than the keyword of “voyage” that might look and maybe act a little bit more? The way the powers that be say that okay, this stuff may say like how are you actually truly determining what keyword is going to be the most value and get you the best ROI? I use Jonas’s example so I’m going to call on him with the voyage and cruise.

Jonas Sickler  

There’s a lot of trial and error. There’s a lot of looking back and analysis. Track everything. Sometimes you’re going to have people coming in from all different points because they’ve run into competitors, or they’ve run into affiliates, or content aggregators along the path as they’ve compared one brand versus another.

Everyone’s not going to take the same path to purchase, and so they’re going to interact with content on your website differently. I think the most important thing is to make sure that you don’t have holes in the journey, first of all, so that you are able to have that big net as Jeremy said earlier for when people do come in.

But you also need to understand the sequence of the journey.

What is the likely next step?

So if they do happen to have their first encounter on your website with something that is deeper in the funnel, what’s the next step that you want them to take?

Have CTAs to that content and measure it and understand that if something’s not performing, tweak the language for the CTA. If there seems to be sort of this natural pronging off where people are visiting a different page, even though you didn’t build it into your funnel, maybe there’s a reason for that. And then go back and inspect the journey and watch user videos, look at heatmap data. So, you’re not just looking at “Well, we rank for these keywords, and we have all of them on our website.”

But are you knitting those into the journey somehow? You know, is the journey that you’ve orchestrated the one that people naturally want to take? And if not, go back and optimize that journey.

Paul Shapiro  

Yeah, a lot of this will depend on how much data you have and what you do with it. Like Shopify has every keyword match to a conversion rate and what position in the funnel it is, and we can project exactly what kind and how many conversions we’re going to get at each keyword. It’s something that moves up and down and weighs things accordingly. That’s, that’s great. That’s not always easy for depending on, like the size of a business. I think that’s probably an ideal state.

But I also want to caution some people. I think, sometimes, like SEO gets a little myopic with the keyword lists, and we don’t talk enough about the full picture. So, when you start thinking about things through more of a product lens versus a marketing lens, that completes the full picture.

So, it’s not really just one or the other — marketing or product work — but you need the full picture. And he did try to break that down, like marketing and SEO. That’s, you know, how are people going to get to your website, to your content, to your feature, and then to your product to sort of complete and bring it full circle? Yeah.

Jeff Coyle  

Jeremy, you and I are on it today. I’ll just highlight what Paul said resonates really significantly, because it’s about what data you have access to. I would also add another caution to Paul’s caution. The caution I would say is that a lot of people are focused on telling you what you can write to win with something Paul mentioned. Telling you that you can’t win without off page, or you can’t win without certain things.

You need the breadth and depth of coverage. You need high quality content that’s exhibiting expertise. You do need to understand your off page, your historical success rates, and your historical authority, both from a link perspective and your existing momentum to be able to make the decisions about how much content you need in order to maintain or grow.

Don’t allow yourself to read a font of some sort saying it’s all about this. It’s all about links. It’s not all about one thing. If it was all about one thing, then you wouldn’t have people like Paul managing a massive BI platform for predictive analysis. You know, I sell the six-billion-dollar keyword database to, you know, many large publishers who build out these big BI systems to try to predict this.

You can get a little bit predictive. You just understand your existing presence, your momentum, and your off-page factors. We don’t think you can’t win just because you’re not as big as your competitors. Because you can.

Jonas Sickler  

On top of that, too, I think there’s definitely a tendency for SEOs to kind of pile onto the whole, you know, “head term has high volume” mindset. We recently did an analysis where we mapped out all of the content on a blog for one of our customers and we bucketed those posts out, and we saw that the top three posts drove 38% of the conversions. THREE. And then you look at the next 10, and 19% of the conversions came for the next ten.

Then 43% of the sitewide conversions came from the next 1,400 pages.

So, all of those pieces of content coming together. And I know Jeff, you’ve got 15 or so articles about topic clusters on your site.

Now most of those may not rank, but they’re part of the story. They may convert if somebody finds them. There’s the zero-keyword, zero search volume, whatever keyword lists, so it’s not really about that, it’s more about, “Are you covering the topic for your entire audience?”

The value that you might get from these really long-tail specific queries could be more than the value you get from the higher volume head terms that people are always focused on and competing for, like the red meat, so to speak.

Jeremy Rivera  

There’s one minute left with you guys. So, 60 seconds on the clock. What’s your final shot?

Paul Shapiro  

Yeah, I mean, I’d say just don’t fall into that trap of being the myopic SEO. Consider the full user journey. Think of your keywords as a portfolio with, you know, each of those keywords contributing a little bit here and a little bit there, creating a safety net and room for opportunity.

Jeff Coyle  

All the content you have on a topic is working together as an asset to enable you to open up new opportunities, as well as perform for those most important money pages, or things like that. So, don’t value pages in isolation.

There’s nothing better than to compete with someone in SEO when all they do is JUST sort their keyword list by descending search volume.

Jonas Sickler  

Think about the overall value of all the assets, all the digital assets that you’re creating together as a lump sum and focus on the controllable things that you can on your website first. Then try to take back some of the market share from the aggregators and the publishers that are encroaching on your market share.

More from the author: