Keyword research used to be the most reliable lever in SEO. Find the phrases people search, assess the competition, build the right page, and you’re off to the races. That logic still holds but the search experience around it has changed fast. In the GEO era (Generative Engine Optimization), your keyword strategy can’t stop at “what will people type into Google?” It has to include “what will people ask in AI-driven interfaces, and what sources will those systems choose to summarize?”
That distinction matters because the reward structure is shifting. Some queries still behave like classic SEO: the user searches, scans results, clicks, and converts. But many queries now end earlier—inside the SERP or inside a generated answer, where the user might get what they need without ever visiting a website. The result is a weird feeling many marketers have right now: rankings might be stable, impressions might even be up, but clicks are softer than they used to be. Keyword research didn’t break. The surface area expanded.
The good news is that keyword research is actually more valuable in this environment and not less. It’s just that the output has to evolve. A modern keyword strategy isn’t a spreadsheet of terms and volumes. It’s a content roadmap that accounts for buyer intent, SERP layouts, and the kinds of content that get selected as “the answer.” If you get that right, you don’t just rank. You become the reference.
What GEO changes about “intent”
Old-school keyword research treated intent as a simple funnel: informational, commercial, transactional. The GEO era adds another layer: answerability. Some queries are inherently “clickable” because the user needs depth, options, pricing, or a tool. Others are inherently “answerable” because the user wants a definition, a checklist, a quick comparison, or steps.
That difference determines what kind of content wins.
A query like “Shopify SEO agency” is still click-driven. People want options and proof. But a query like “what is topical authority” is answer-driven. It’s easy to summarize and satisfy quickly—meaning the winner isn’t always the page with the most backlinks. It’s often the page with the clearest definition, the cleanest structure, and the most direct explanation.
So in GEO-era keyword research, you’re not only matching intent. You’re matching the format the SERP (or AI layer) is trying to deliver.
Why “volume + difficulty” on it’s own is a trap now
Search volume and keyword difficulty still matter, but they can be misleading on their own. A high-volume keyword that triggers heavy SERP features (snippets, PAA, AI summaries) may have less click potential than the tools imply. Meanwhile, lower-volume queries often punch above their weight because they’re closer to decision-making, easier to rank for, and more likely to result in a lead.
In practice, the best keyword opportunities today tend to share a few traits:
- They map cleanly to a real business problem or decision
- They have “messy” SERPs (meaning Google is unsure what the best answer format is)
- They reward specificity (templates, examples, frameworks, comparisons)
- They can be expanded into a cluster, not just a one-off post
If your research process doesn’t include those filters, you’ll create a lot of content that looks busy but doesn’t compound.
The new center of gravity: clusters that make you the source
The biggest mistake we see with content in the GEO era is the “one post, one keyword” mentality. That worked when ranking a single page could reliably drive traffic for an entire topic. But in 2025, the sites that consistently show up (both in traditional rankings and in the answer layer) tend to publish coverage, not just content.
Coverage means you don’t publish “SEO keyword research” as one blog post and call it done. You publish a set of connected assets that answer the topic from multiple angles, and you link them together in a way that mirrors how people actually learn and decide.
A strong cluster often includes:
- A main guide that defines the concept and sets context
- A step-by-step process post (how to do it)
- A “mistakes to avoid” post (what fails in real life)
- A comparison post (tools, approaches, in-house vs agency)
- A template/checklist (downloadable or copyable)
- A case study or example (proof it works)
This cluster approach does two things at once: it increases your chance of ranking across a wider set of queries, and it increases the chance your site becomes a source that AI systems reference repeatedly.
What keyword research looks like when you do it like a marketer
Here’s the key shift: keyword research shouldn’t start inside a tool. It should start with how your customers talk when they’re trying to solve a problem.
Instead of just building a seed list of services (“technical SEO,” “content strategy,” “link building”), build a seed list of moments:
- “Our traffic dropped after a redesign”
- “We migrated platforms and lost rankings”
- “We’re publishing content but it’s not converting”
- “We’re growing but SEO is stalled”
- “We need to prove SEO impact to leadership”
- “We’re being outranked by worse competitors”
- “We’re seeing impressions but no clicks”
Each of those moments expands naturally into keyword clusters that are both high-intent and GEO-friendly. And they lead to content that reads like expertise instead of generic SEO tips.
From there, you can expand using the normal channels: Search Console queries, Ahrefs/Semrush suggestions, PAA questions, competitor top pages, and internal sales call notes. The difference is what you do next: you categorize and prioritize based on what’s likely to win in today’s SERP environment.
The content formats that win in GEO-era keyword research
If you want your content to perform in a world where answers are often generated, you need to publish in formats that are easy to extract and hard to misunderstand. That doesn’t mean writing for robots. It means writing with clarity.
Here are formats that consistently outperform “standard blog posts”:
- Definitions with nuance: short, clear, and accurate and then expanded with context
- Step-by-step workflows: numbered steps, decision points, “if this, then that” logic
- Checklists and templates: especially for audits, migrations, refreshes, tracking
- Comparisons: tools, approaches, tradeoffs, when to use what
- Examples: screenshots, mini case studies, “good vs bad” patterns
- Frameworks: simple mental models that decision-makers can repeat internally
This is also why a lot of “thin” content is getting crushed right now. It’s not that the writing is bad. It’s that the format doesn’t provide anything a system can confidently reuse.
How to prioritize keywords when clicks are less guaranteed
One of the most useful shifts you can make is to prioritize keywords using four factors instead of two:
- Business value: will this attract a buyer or an influencer?
- Ranking feasibility: can we realistically win within our authority level?
- SERP click potential: does this query still produce clicks?
- Source potential: does this query produce citations/summaries?
When you do this, you naturally stop chasing vanity topics and start building assets that compound. You’ll also notice something: the best topics are often the ones where you can bring your agency’s real-world experience to the surface, because that’s what differentiates you from the 500 generic posts already out there.
The main takeaway
SEO keyword research in the GEO era is still keyword research, but the finish line moved. You’re not only trying to rank a page. You’re trying to become the source that gets referenced when the topic comes up, whether the user clicks or not.
That’s why the best keyword research today produces a plan that includes intent, format, and coverage and not just terms. The sites that win will be the ones that treat content like product: structured, useful, and obviously written by people who do this work every day.