The last brief I killed was 2,400 words of outline for a plumbing service page. Solid H2s. Sensible FAQ. Real keyword research underneath it. I deleted 40% of it before the writer opened the doc.
The reason is boring and it is the reason most content briefs produce filler. Every section had a heading. Not every section had anything only that client could say.
I run a test on every brief before it goes to a writer. It has nothing to do with word count, keyword density, or how the outline reads. It has one question. Could you paste this section onto a competitor's page for the same keyword and would anyone notice.
If the answer is no, the section stays. If the answer is yes, I cut it. Usually 30 to 40% of a first-draft brief fails.
Why outlines pass and depth fails
A good outline is a map of what the SERP wants. You read the top ranking pages, you classify format, you note the sections that keep appearing, you build a structure that matches. That work is real and it is necessary. But it produces a page that looks like the SERP average, not a page that beats it.
Google is looking at the same SERP you are. It has already seen the average of these sections a thousand times. What it does not have is your client's actual process, actual pricing structure, actual crew size, actual objections from actual buyers.
The outline is the shell. The depth is the animal inside it. Most agencies ship the shell empty and then complain that content marketing does not work.
What the test actually asks
Before a brief goes to a writer, I go section by section with 3 questions.
| Question | What passing looks like | What failing looks like |
|---|---|---|
| Could this section paste onto a competitor's page without anyone noticing? | No, it names something specific to this business (a technique, a service constraint, a real objection) | Yes, it is a general explanation of the topic anyone could write |
| Does the writer need to interview or observe the operator to write this? | Yes, they need input the client alone has | No, they can pull it from the first 3 SERP results |
| Would a buyer who has already read 4 competitor pages learn something here? | Yes, they see something new that reframes the decision | No, they get the same overview they have read 4 times |
Any section that fails 2 of 3 gets cut. Any section that fails all 3 gets deleted without discussion. Anything that fails just 1 gets flagged for the writer with a specific research task, usually a 15 minute call with the client to fill the gap.
What this looked like on a real page
The Tenmar page is the cleanest example I can point to. Glass railing installer in BC. When we started, they had 5 pages and were losing every comparison search to national aggregators.
The obvious play was to build 80 new pages fast, each targeting a different service and location combination. Standard hub-and-spoke. Any agency reading this could have drawn the same map.
What made it work was not the map. It was that every page had a section a competitor page did not. On the frameless railing page, we included the load-bearing math the crew actually uses on site, translated for a homeowner. On the location pages, we named the specific municipal permit quirks that catch homeowners out. On the comparison pages, we ran the numbers a competitor would not put in writing because it made their higher-margin service look worse.
85 pages in 3 weeks. 4x lead generation in 10 weeks. The scale mattered, but the pages ranked because each one had something you could only write if you had spent time with the crew. The full Tenmar breakdown is on the case study page.
The interview question I ask every operator
To find the sections a writer alone cannot produce, I ask the client one question on every kickoff.
What do you say to a customer on the phone that you have never seen written on your website. Usually there are 5 to 10 of these. Objections handled a specific way. Process details the client thinks are obvious. Price-tier logic. Reasons they turn work away.
Every single one of those becomes a section a competitor cannot copy. Because to copy it, they would have to know how your business actually runs. The test I run on their answers is the same test I run on the brief. If the answer sounds like something any operator in the category would say, it is not depth, it is filler dressed as insight. I keep pushing until I hear something that would make a competitor uncomfortable to see in writing.
Why this matters more in 2026 than it did in 2022
The gap between a shell page and a depth page used to be a ranking gap. Now it is a ranking gap and a citation gap.
25% of Google searches now show an AI Overview, up from 13% six months earlier (Conductor, 2026). Traffic from AI tools to websites grew 500% through 2025 (Previsible, 2025). AI systems cite pages that say something specific. They skip pages that say the same thing every other page in the category says, because there is no reason to pick one over another.
Across dozens of clients, the pattern is the same. The pages that get pulled into AI Overviews, ChatGPT answers, and Perplexity citations are the pages with sections a competitor could not write. AetherHaus picked up 89 AI citations across 5 different AI systems because the pages named specific protocols, specific temperatures, specific recovery outcomes, not because we added FAQ schema.
What to do this week
Pull your 3 most important service pages. Read each section out loud. For each one, ask whether you could paste it onto a direct competitor's website and change nothing. Everything that could be pasted comes out. Everything that stays gets rewritten with something only your business could say.
You will lose word count. You will lose the reassurance of a page that looks the same length as the ones that already rank. What you will gain is the only thing that actually beats the SERP average, which is not being the average.
The format test tells you what shape the page should be. The depth test tells you what belongs inside it. One without the other produces a page that either looks wrong or reads like everyone else. Both together produce a page that ranks and gets cited.
Liam Lytton is the founder of The 66th, a Vancouver-based SEO and GEO agency. He works directly with founders on every account.
