Last year a prospect sent me their content plan before our first call. 47 keywords, all green-lit by their previous agency, all tagged "high commercial intent." One of them was "best cold plunge for recovery." Their page was a 600-word service page selling installation in Vancouver.
The top 5 results were all listicles. Wirecutter, Garage Gym Reviews, two affiliate sites, one Reddit thread. Zero installer pages. Zero local pages. The keyword had volume. The intent was research. The page was a transaction. No amount of on-page work was going to close that gap.
This is the most expensive mistake I see in SEO, and it has nothing to do with backlinks, schema, or technical depth. It is a format mismatch, and you can catch it in 90 seconds before you commit a single hour of writing.
Format match beats page authority, every time
You can have the strongest domain on the internet. If your page format does not match what is winning page 1, you are fighting gravity. Google has tested every URL on page 1 against millions of real searchers. The format that ranks is the format that wins. Your instinct about what the page "should" be is what you would want. It is rarely what the searcher wants.
Before any keyword enters one of our content plans, I open the actual ranking URLs and classify each one. I am looking for one thing: do these pages agree on what they are? If 4 of the top 5 are comparison listicles and you are writing a service page, the keyword does not belong on your service page. It belongs on a comparison page, or it does not belong on your site at all.
The 90-second test, step by step
Here is exactly what I do, in order, for every keyword that survives the basic relevance check.
1. Open the top 10 results in 10 tabs. Not a tool. The actual pages. Tools give you a category guess. I want my own classification.
2. Classify each URL into one of 5 buckets. Service page, listicle or comparison, blog or guide, product or category page, or directory or aggregator. I write the bucket next to each URL in a sheet.
3. Count the majority. If 6 or more of the top 10 are the same format, that is the format Google has decided this keyword wants. If it is split 4-3-3, the keyword has mixed intent and is a different conversation.
4. Match or move on. Either your page is the majority format, or you build a new page in that format, or you drop the keyword. There is no fourth option. "Optimising our existing page harder" is not a strategy when the format is wrong.
What the buckets actually mean
The 5 formats are not equal. Each one is telling you something different about who is searching and what they want.
| Format winning page 1 | What the searcher actually wants | Where your page belongs |
|---|---|---|
| Service or location page | Ready to hire or buy now | BOF service page on your site |
| Listicle or comparison | Shortlisting before deciding | MOF comparison page, separate URL |
| Blog or guide | Learning, not yet buying | TOF blog post, separate URL |
| Product or category page | Browsing inventory | Ecommerce category, not a service page |
| Directory or aggregator | Trusts a third-party list more than a brand | Get listed on the directory, build your own listicle, or skip |
The most common mistake I see is collapsing the middle 3 into "content." They are not the same. A comparison page and a blog post sit at different points in the buyer's journey, and trying to make one page do both is how you end up ranking for neither.
The cold plunge story, finished
I will not name the prospect, but a similar situation played out with a wellness client we eventually took on. Their team had been writing service-style pages for keywords that were unambiguously research keywords. The top 10 were listicles. Their page was a 700-word transactional layout.
The fix was not on that page. The fix was building a separate comparison page for the research keyword, leaving the service page focused on its own buyer search (the local, location-included version), and letting each page serve the format Google had already validated. Inside 3 months, the service page hit #1 for its real target. Inside 6, the comparison page was earning AI citations across the major engines. The same site that had been stuck for a year was now compounding.
This is what the pyramid framework actually looks like in practice. You build full-funnel content because the funnel is real, and each page only earns rankings when its format matches the stage of the buyer it is meant to serve. Our SEO services page exists to convert one specific search. Our blog exists to win a completely different set of searches. Neither page does the other's job, and that is by design.
Sourced data that backs this up
Semrush's analysis of SERP feature data has shown for several years that pages matching the dominant content type for a query rank materially higher on average than mismatched pages, even when authority metrics favour the mismatched page (Semrush, 2024). The mechanism is straightforward. Google's ranking signals include click satisfaction patterns from earlier searchers, and a format mismatch produces a measurable signal: people click, bounce back, and click a different result. That signal compounds against you every month the wrong page is live.
What this changes about your content plan
Volume is the least useful column in keyword research. A keyword with 50 monthly searches where the SERP is dominated by service pages is more valuable to a service business than a keyword with 5,000 searches where the SERP is dominated by Wirecutter and Reddit. The first one you can win. The second one you cannot, no matter how much you spend.
If your current content plan has not classified the format of each target keyword's SERP, it is not a plan. It is a list. Across dozens of audits The 66th has run, the single most common reason a site is stuck despite "doing everything right" is a content plan full of keywords whose ranking format does not match the page being optimised.
3 questions to put against every keyword before it makes the cut:
Does the format of the top 6 results match the page we plan to put this on? Would a buyer landing on our page from this search feel they found the right kind of result? If we built a new page in the winning format, would it conflict with a page we already have?
If you cannot answer all 3, the keyword does not belong in the plan yet. Move it back to research and run the test.
Liam Lytton is the founder of The 66th, a Vancouver SEO and GEO agency working with founder-led businesses across Canada, the US, and the UK. He writes about what actually moves rankings, not what sounds clever in a deck.
