Most brands still treat keyword research like a volume report. They export a list from a tool, sort by search volume, pick the biggest numbers they think they can win, and hand the list off. Then they wonder why ranking pages do not convert and why the articles they published never built real authority.
That workflow was already breaking in 2022. In 2026, with AI Overviews appearing on a large share of informational queries and search engines reading entities instead of strings, it does not work at all. Modern keyword research is less about finding big numbers and more about mapping what people actually want, which pages should earn the clicks, and how each keyword fits into a cluster your brand can legitimately own.
This guide covers how we approach keyword research at EmberTribe for DTC brands and growth-stage SaaS companies, and the mistakes we see burning budget on content and paid search.
The short version: keyword research is the process of discovering the queries your potential customers type, ask, or prompt, then understanding the intent behind each one well enough to decide what type of asset should answer it.
The old definition stopped at "find search terms with good volume and low difficulty." The new definition has to account for four shifts:
Keyword research that ignores any of these produces the same thing it always did: a spreadsheet with big numbers and no plan.
The biggest upgrade you can make to your keyword research process is to lead with intent and treat volume as a tiebreaker, not a filter.
Every query sits in one of four traditional intent buckets: informational, navigational, commercial investigation, or transactional. In 2026, that classification is not granular enough. The pages that win now match narrower sub-intents, things like comparative, instructional, reassurance, and problem-solving intent, each of which calls for a different content format.
A keyword like "best running shoes for flat feet" looks transactional on the surface. Look at the SERP and you see listicles, shoe brand category pages, and a People Also Ask block full of medical questions. The real intent is comparative and reassurance-driven, so a product page will not win that query. A comparison guide built around pain points will.
The practical workflow we use:
This is slower than sorting a CSV. It stops you from chasing terms you cannot rank for, and it tells you exactly what kind of page to build.
General SEO advice falls apart fast when applied to an ecommerce catalog. Ecommerce brands do not need one keyword per post. They need an architecture that maps collections, products, and content to different layers of demand.
As an ecommerce SEO consultant, the first thing we do with a new DTC client is separate their keyword universe into three jobs:
Collection page queries. These are your category-level commercial terms, things like "merino wool base layers" or "leather crossbody bags." They have the broadest commercial intent and drive the most organic revenue per page. Each collection page should own one primary keyword and three to five secondary terms, with supporting content cleaned up so the collection is the clear canonical answer.
Product page queries. These are the narrower, often long-tail terms that signal a shopper near the bottom of the funnel. "Smartwool 250 base layer men's medium" converts at rates a generic category page cannot touch. Most brands underinvest here because the volume looks small, even though revenue per click is the highest in the catalog.
Informational queries. These are the upper-funnel questions, buying guides, and problem-led searches that feed category pages with topical authority. They rarely convert directly. They exist to help collections rank and to earn citations in AI answer engines. This is where most brands working with ecommerce seo companies fall short: they either skip informational content entirely, or publish it in isolation with no link path to the commercial pages.
The mistake we see most often is treating every keyword as equally valid for any page type. If product pages target the same terms as collection pages, you are competing with yourself. If blog content is not explicitly feeding topical authority into your collections, it is a cost center pretending to be ecommerce content marketing.
Clustering is where intent work turns into a content plan. A good cluster is a small set of related keywords that share a primary intent and can be answered by one page well enough to compete. A bad cluster is a dumping ground for anything that shares a noun.
Our rule of thumb: if you can write one honest answer that satisfies every keyword in the group without contradicting itself, it is a cluster. If you cannot, split it.
Inside a cluster, one keyword is the anchor. That is the term that drives the URL, the H1, and the canonical intent of the page. The rest are secondary terms you weave into H2s, FAQs, and body copy. This matches how search engines actually read pages in 2026, where entity relationships and semantic context matter more than exact-match keyword density.
Across a site, clusters roll up into pillars. A pillar is a broad topic your brand wants to be known for, supported by five to twenty interlinked cluster pages. That is how topical authority gets built, and why one-off posts rarely move rankings anymore.
For a SaaS company, a pillar might be "product-led growth" with clusters for activation metrics, freemium models, onboarding flows, and expansion revenue. We walk through how this shows up in practice inside our complete guide to SaaS SEO, and it is one of the specific things to ask about when you are vetting a SaaS SEO agency.
One quiet upgrade modern keyword research makes possible is using the same work to brief both SEO and paid search teams. They are usually treated as separate workstreams with separate keyword lists. That is wasted effort, and it creates inconsistent messaging across the funnel.
The paid team cares about commercial intent, cost per click, and conversion rate. The SEO team cares about volume, difficulty, and topical fit. Intent-tagged, clustered keyword research gives both teams what they need from one source of truth.
A few patterns we use consistently:
One well-built keyword map can inform ad group structure, negative keyword lists, ad copy angles, and a content calendar at the same time.
There is no shortage of keyword research tools. The honest answer is that most of the best data comes from combining two or three, not from buying the most expensive all-in-one platform.
The tool matters less than the workflow. A disciplined researcher with Search Console and the actual SERPs will beat a sloppy operator with a five-figure SaaS stack.
The mistakes we see most often when we audit a brand's keyword strategy:
If your current keyword research is a spreadsheet sorted by volume, start over. Pull your list, open the SERPs, tag intent, and regroup everything into clusters mapped to the page type that should own each group. That single exercise is usually worth more than a new tool subscription.
From there, decide which two or three clusters your brand can legitimately own in the next two quarters, map them to specific collection pages, product pages, or content hubs, and use the same research to sharpen your paid search targeting. The payoff is a keyword strategy that pulls its weight in both channels instead of living in isolation on a strategist's laptop.
EmberTribe runs keyword research as part of every integrated paid media and SEO engagement for DTC and SaaS clients, which means the work never sits on a shelf. If your keyword strategy feels more like a list than a plan, we can help you rebuild it around intent and clusters that hold up in AI search.
Most brands still treat keyword research like a volume report. They export a list from a tool, sort by search volume, pick the biggest numbers they think they can win, and hand the list off. Then they wonder why ranking pages do not convert and why the articles they published never built real authority.
That workflow was already breaking in 2022. In 2026, with AI Overviews appearing on a large share of informational queries and search engines reading entities instead of strings, it does not work at all. Modern keyword research is less about finding big numbers and more about mapping what people actually want, which pages should earn the clicks, and how each keyword fits into a cluster your brand can legitimately own.
This guide covers how we approach keyword research at EmberTribe for DTC brands and growth-stage SaaS companies, and the mistakes we see burning budget on content and paid search.
The short version: keyword research is the process of discovering the queries your potential customers type, ask, or prompt, then understanding the intent behind each one well enough to decide what type of asset should answer it.
The old definition stopped at "find search terms with good volume and low difficulty." The new definition has to account for four shifts:
Keyword research that ignores any of these produces the same thing it always did: a spreadsheet with big numbers and no plan.
The biggest upgrade you can make to your keyword research process is to lead with intent and treat volume as a tiebreaker, not a filter.
Every query sits in one of four traditional intent buckets: informational, navigational, commercial investigation, or transactional. In 2026, that classification is not granular enough. The pages that win now match narrower sub-intents, things like comparative, instructional, reassurance, and problem-solving intent, each of which calls for a different content format.
A keyword like "best running shoes for flat feet" looks transactional on the surface. Look at the SERP and you see listicles, shoe brand category pages, and a People Also Ask block full of medical questions. The real intent is comparative and reassurance-driven, so a product page will not win that query. A comparison guide built around pain points will.
The practical workflow we use:
This is slower than sorting a CSV. It stops you from chasing terms you cannot rank for, and it tells you exactly what kind of page to build.
General SEO advice falls apart fast when applied to an ecommerce catalog. Ecommerce brands do not need one keyword per post. They need an architecture that maps collections, products, and content to different layers of demand.
As an ecommerce SEO consultant, the first thing we do with a new DTC client is separate their keyword universe into three jobs:
Collection page queries. These are your category-level commercial terms, things like "merino wool base layers" or "leather crossbody bags." They have the broadest commercial intent and drive the most organic revenue per page. Each collection page should own one primary keyword and three to five secondary terms, with supporting content cleaned up so the collection is the clear canonical answer.
Product page queries. These are the narrower, often long-tail terms that signal a shopper near the bottom of the funnel. "Smartwool 250 base layer men's medium" converts at rates a generic category page cannot touch. Most brands underinvest here because the volume looks small, even though revenue per click is the highest in the catalog.
Informational queries. These are the upper-funnel questions, buying guides, and problem-led searches that feed category pages with topical authority. They rarely convert directly. They exist to help collections rank and to earn citations in AI answer engines. This is where most brands working with ecommerce seo companies fall short: they either skip informational content entirely, or publish it in isolation with no link path to the commercial pages.
The mistake we see most often is treating every keyword as equally valid for any page type. If product pages target the same terms as collection pages, you are competing with yourself. If blog content is not explicitly feeding topical authority into your collections, it is a cost center pretending to be ecommerce content marketing.
Clustering is where intent work turns into a content plan. A good cluster is a small set of related keywords that share a primary intent and can be answered by one page well enough to compete. A bad cluster is a dumping ground for anything that shares a noun.
Our rule of thumb: if you can write one honest answer that satisfies every keyword in the group without contradicting itself, it is a cluster. If you cannot, split it.
Inside a cluster, one keyword is the anchor. That is the term that drives the URL, the H1, and the canonical intent of the page. The rest are secondary terms you weave into H2s, FAQs, and body copy. This matches how search engines actually read pages in 2026, where entity relationships and semantic context matter more than exact-match keyword density.
Across a site, clusters roll up into pillars. A pillar is a broad topic your brand wants to be known for, supported by five to twenty interlinked cluster pages. That is how topical authority gets built, and why one-off posts rarely move rankings anymore.
For a SaaS company, a pillar might be "product-led growth" with clusters for activation metrics, freemium models, onboarding flows, and expansion revenue. We walk through how this shows up in practice inside our complete guide to SaaS SEO, and it is one of the specific things to ask about when you are vetting a SaaS SEO agency.
One quiet upgrade modern keyword research makes possible is using the same work to brief both SEO and paid search teams. They are usually treated as separate workstreams with separate keyword lists. That is wasted effort, and it creates inconsistent messaging across the funnel.
The paid team cares about commercial intent, cost per click, and conversion rate. The SEO team cares about volume, difficulty, and topical fit. Intent-tagged, clustered keyword research gives both teams what they need from one source of truth.
A few patterns we use consistently:
One well-built keyword map can inform ad group structure, negative keyword lists, ad copy angles, and a content calendar at the same time.
There is no shortage of keyword research tools. The honest answer is that most of the best data comes from combining two or three, not from buying the most expensive all-in-one platform.
The tool matters less than the workflow. A disciplined researcher with Search Console and the actual SERPs will beat a sloppy operator with a five-figure SaaS stack.
The mistakes we see most often when we audit a brand's keyword strategy:
If your current keyword research is a spreadsheet sorted by volume, start over. Pull your list, open the SERPs, tag intent, and regroup everything into clusters mapped to the page type that should own each group. That single exercise is usually worth more than a new tool subscription.
From there, decide which two or three clusters your brand can legitimately own in the next two quarters, map them to specific collection pages, product pages, or content hubs, and use the same research to sharpen your paid search targeting. The payoff is a keyword strategy that pulls its weight in both channels instead of living in isolation on a strategist's laptop.
EmberTribe runs keyword research as part of every integrated paid media and SEO engagement for DTC and SaaS clients, which means the work never sits on a shelf. If your keyword strategy feels more like a list than a plan, we can help you rebuild it around intent and clusters that hold up in AI search.
When am I going to start seeing results?
How fast can we scale to $25,000?
How much am I going to spend on testing?
These questions (and more) come up frequently as we're talking to companies who are considering working with us to grow their business. Whether they are just starting out on a new eCommerce store or looking to increase their app signups 3x in Q2, the underlying question is really the same.
Let's face it: digital marketers (and marketing agencies) have really turned their approach into a "black box" over the years. Whether they do it by hiding behind jargon, slapping clever branding over the top, or creating complex or confusing diagrams, the end result is confused business owners who don't really understand what their dollars are going towards, or why.
Now, take a deep breath.
For you, with us, that stops here.
We're about to open the box.
I have two little kids, one preschooler and one toddler. Both are (alarmingly) ambulatory, moving all over the house and getting into everything they aren't supposed to. The older one can unlock deadbolts, push open screen doors, climb ladders and stairs, while the younger is content with simple seeing how fast she can get her body moving in a single direction before she either topples forward or encounters an object that refuses to budge when she slams into it.
Why do I bring this up? Because they didn't start this way.
Yes, it's a tired cliche, but it's so true: you have to walk before you can run.
If your business has never run an ad before, never used marketing to sell, never attempted to convince someone unfamiliar with the brand, product, or service that they should part with their hard-earned Benjamin Franklins, then your first question should not be, "How much can I make?"
You don't have TRACTION yet.
By traction I mean a pattern of desired behavior occurring in a consistent, somewhat predictable fashion. This could mean generating leads, getting purchases, onboarding new users or whatever else your business goal, it doesn't matter. The point is that you need to be able to say that you can cause it to happen, repeatedly, with your efforts.
When we work with clients who have never run ads before, or who are just starting out, our first forays out into the marketplace are focused on finding who will buy and what will cause them to buy. Put another way, this is about audience and creative/offer.
Let's bust a myth: just because you have a product or service does not mean people will buy it. This is not Field of Dreams.
On the contrary, you have to wade through scores of unqualified or uninterested people to find your best candidates, and then test multiple different messages, angles, images, videos, taglines and more in order to find traction.
"Okay, but how long does that take?"
Well, that depends.
I know, that's not what you were hoping for. And if I can tell you that it would take 2 weeks or 2 months or whatever, I definitely would. Instead, here's what I can tell you:
When you work with us, you aren't hiring wizards (or gurus or ninjas) - you're hiring data-driven marketers. So we're going to test, and test, and test, and generate lots of data, and then we're going to do what the data tells us.
👉 Set up your campaigns to get more qualified leads. →
Sometimes it's fast, and we see traction in just a few weeks. Sometimes it takes less time, sometimes it takes longer. All the factors above impact that.
But the good news is, once you have TRACTION, you can move on to start thinking about...
Too many times we'll talk to a business owner who is putting money into ads and wants to see immediate return. If they don't get a certain CPA or ROAS in the first 3 weeks, they think there's something "wrong" with the ads. They don't realize they are trying to run before walking, that you can't build a house without the foundation, or whatever analogy you best identify with.
💡 ROAS isn't everything, it's just a part of the equation. →
Once we help our clients find TRACTION, then (and only then) is it time to start discussing PROFIT.
Why?
If you don't have enough data points, you can't optimize.
Put another way, if you don't have anyone buying from you, how do you know who your best customers are?
Getting this data and acting on it is the basis of improving your PROFIT metrics. If you want a better CPA, you need to find out which creative gets the best response and then test small optimizations on it - a new emoji, a different headline, a carousel vs a static image. If you want better ROAS, you can segment by device type or placement or time of day that gives you the best baseline.
The key to the PROFIT stage is having goals. And I don't mean "I want to retire and sleep on a bed of Andrew Jacksons every night" type goals, more like "If I can generate new users for $20 each that means I'm profitable and am basically printing money" goals.
We help our clients walk through some simple calculations to set their goals. For an eCommerce store this might include repeat purchase rate, average order value (AOV), and cost of goods sold (COGS). For a SaaS client, we would consider lifetime value (LTV), profit margins, and upsells. Whatever the case, we want to end up with a single number.
That number is our PROFIT goal. If we can hit that goal with consistency, it unlocks us to move on the third and final stage.
Ah yes, scaling. The magical, mystical land of unicorns and rainbows where you trade $1 for $4 ten thousand times while eating ice cream in your pajamas.
Okay, well, not quite, but that's how the "get rich quick" YouTube personalities pitch it. Sounds fun, huh?
Truth is, scaling isn't the end - it's the beginning.
When this client partnered with EmberTribe, their goal was to find strategies to scale sales. Now our client has experienced scale from $18K to $370K lifetime revenue, with an $111K lifetime spend.
You can't start putting more dollars into your campaigns until they are making you money back consistently, and you can't do that until you build a system of repeatable client generation. Hence the reason it's the final step. But there's another reason we counsel clients to be smart about getting to this stage: the game changes.
If you want to triple your investment in ads, especially on a channel like Facebook Ads, just about the worst thing you can do is start jacking your budgets up quickly.
😑 Facebook ads not working? This could be why. →
This causes the algorithm to have to start relearning, and oftentimes can tank your PROFIT, forcing you to go back to the drawing board. Instead, you have to be intentional, constantly revisiting your PROFIT goals and testing new TRACTION experiments to widen your funnel. And this is why we insist on walking through the process with clients - because failing to do the hard work on the front end ends up in a house of cards that falls apart, leaving everyone unhappy.
Some of our best success stories are clients who did things the right way, worked with us to build a repeatable system for growth and testing, and then let us run wild with new audiences, creative, automation, rules and more. Their accounts grew from 5 to 7 figures in ad spend profitably not by some mystical proprietary technology or the wizardry of a paid acquisition savant, but by being intentional, creating a solid foundation, and trusting the process.
It's not easy. It's not as fast as we'd like. But the results are worth it, and the potential that it opens up are amazing.
No black box. No magic. No single genius with the inside track on the algorithm.
Just lots of testing, patience, observation, analysis, failure, growth and consistency.
That's the secret sauce of EmberTribe, and it's one of the reasons we've had such great success for ourselves and our clients since our inception: a three-step process of TRACTION, PROFIT, SCALE that works across industries, across business models, regardless of the age or success of the business to date.
.svg){kind=icon}