The SEO Boardroom
Claude agrees with whatever you want to hear. The SEO Boardroom runs your SEO idea past five adversarial advisors plus a Chair, who hunt for where it fails, disagree with each other, then hand down one verdict: ship, fix first, or kill.
Built by Lawrence Hitches, Chief of Staff at StudioHawk. Grounded in a real SEO framework, not invented.
What it is
The SEO Boardroom is a Claude skill. You bring one real SEO decision, an idea, a content plan, a page you want to build. Five advisors each attack it from one discipline, looking only for where it breaks. Then a Chair synthesises them into a single call and the next step.
The point is disagreement. Ask Claude "is this a good idea?" and it will usually find a way to say yes. Force five separate lenses to hunt for failure, and the bad ideas surface before you spend a month building them. When two advisors flag the same flaw independently, that is not a coincidence, it is the signal.
It is modelled on the multi-advisor "LLM Council" format, but it is SEO-specific and every rule traces to a real framework, built from Google ranking systems, patents, the Quality Rater Guidelines, information gain research, topical authority and AI search optimisation.
Who sits on the board
The Crawler
Can Google and AI bots access, render and index this? Flags JS-only rendering, crawl budget, robots and canonicals, page weight, site health.
The Editor
Does this deserve to rank, or is it a rehash? Flags thin or derivative content, no first-hand data, failing the find-and-replace and regeneration tests.
The Intent Reader
Does this match what searchers want and what the SERP rewards right now? Flags intent mismatch, wrong page type, ignoring what ranks today.
The Authority Check
Do you have the authority to win this, or are you punching above your weight? Flags thin topical coverage, weak entity, going too broad too soon.
The Answer Engine
Will ChatGPT, Perplexity, Gemini and AI Overviews actually cite this? Flags content that is not answer-shaped, no citable passages, no distinctiveness.
The verdict
Synthesises the five: where they agree, where they clash and how it resolves, the blind spots they caught, then one verdict and one next step. No hedging.
How it works
- You bring a real SEO decision, idea or plan. If it is vague, the board asks one question before convening.
- Each advisor attacks it from its lens, grounded in the framework, and returns its top flaws plus a verdict: ship, fix first, or kill.
- The advisors peer-review each other's calls, anonymised, to catch groupthink and overreach. Weak flags get withdrawn here.
- The Chair synthesises one verdict plus the single next step.
A real verdict (nothing invented)
Here is the Boardroom run on a genuine decision from my own site: should I build a page to rank for the head term "what is ai search"? The demand is real and rising, so it is tempting. This is the honest output, including where the board disagreed with itself.
- The Intent Reader and The Authority Check, independently: the live SERP for "what is ai search" is owned by enterprise-search vendors (IBM, Coveo, Sitecore, Microsoft Azure, Databricks). Google reads "AI search" as an enterprise software product, not the answer engines I write about. Different meaning, different result set. And the site is absent from that SERP, because Google reads a bigger entity as the authority there.
- The Editor agreed from a third angle: a plain "what is ai search" definition fails the find-and-replace test. IBM and Cloudflare already answer it. No information gain, so nothing to rank.
- The Crawler said ship (nothing technical stops the page) but withdrew it under peer review: technically possible is not the same as worth doing, and the constraint here was never technical.
- The Answer Engine said fix first, not kill. One thing on this topic is genuinely citable: nobody in that SERP disambiguates the two meanings of "AI search" (enterprise product vs answer engine). That disambiguation is real information gain and answer-shaped, so it can earn citations even where the blue link cannot be won.
- Resolution: do not build a new page for the bare head term. Fold an answer-first disambiguation into the existing hub, and chase the rising, winnable variants ("ai search optimization", "best ai search engine") instead of the term IBM owns.
- The original idea treated "what is ai search" as one query. It is two queries wearing one phrase. Chasing it head-on means competing with IBM for a meaning that is not even mine.
Do not build a page to rank for the head term. Own the answer-engine meaning through disambiguation on the existing hub, and target the rising variants where the authority is winnable. Next step: add an answer-first "what is AI search" disambiguation section to the /ai-search/ hub and an "AI search optimization" section for the rising variant. (That is exactly what shipped, and the board's verdict matched the call.)
Install the SEO Boardroom
- Download the skill. Grab the seo-boardroom.md file.
- Open Customise skills in Claude. Settings, then Capabilities, then Skills. Upload the file.
- Run it. Type
/seo-boardroom, or paste a real SEO decision and say "run the SEO boardroom on this".
FAQ
What is the SEO Boardroom?
A free Claude skill that takes one SEO idea, decision or content plan and runs it past five adversarial advisors plus a Chair. Each advisor owns one discipline and hunts for where the idea fails. They disagree, peer-review each other, then the Chair gives one verdict: ship, fix first, or kill, plus the next step.
Why not just ask Claude directly?
Claude tends to agree with whatever you propose. The Boardroom forces structured disagreement so bad SEO gets caught before it is built. When two advisors flag the same flaw independently, that is the strongest signal it is real.
Is it grounded in real SEO, or made up?
Every advisor rule traces to a real SEO framework built from Google ranking systems, patents, the Quality Rater Guidelines, information gain research, topical authority and AI search optimisation. Advisors cite the rule so you can check it.
When should I not use it?
Not for factual lookups, obvious yes or no questions, or writing tasks. It is built for contested decisions with a real chance of being wrong. A boardroom that rubber-stamps is worse than no boardroom.