Verified against ChatGPT · 2026-08-09
Find the specific subtopics a competitor covers that we don't, ranked by how much it would actually cost us to close each gap
Compares our existing page against a competitor's on the same topic and produces a coverage matrix scored by both SEO opportunity and realistic effort, instead of a flat list of missing subtopics with no sense of what's actually worth doing.
The prompt
Ready to copy — highlighted parts are example details you can swap.
Compare our page against a competitor's page on the same topic and identify content gaps, but score each gap by effort as well as opportunity so this turns into a prioritized list, not just a list. OUR PAGE (paste content or summary) Our page: intro, 'what is a HELOC', 'how to apply', short FAQ (3 questions). No section on rates or tax implications. COMPETITOR PAGE (paste content or summary) Competitor page additionally covers: current HELOC rate ranges by lender, tax deductibility rules, a rate comparison table, and a risk section on variable-rate exposure. TARGET KEYWORD THIS IS COMPETING FOR how does a HELOC work OUR CONSTRAINTS One in-house writer, no licensed financial advisor on staff to review rate/tax claims, two-week timeline for this update. STEP 1 — SUBTOPIC INVENTORY List every distinct subtopic the competitor page covers, and mark which ones our page also covers, partially covers, or doesn't cover at all. Be specific about "partially covers" — a subtopic mentioned in one sentence with no depth is functionally a gap even if the words technically appear on our page, and should be marked as such rather than credited as covered. STEP 2 — SCORE EACH GAP For every subtopic we don't cover or only partially cover, estimate: (a) how directly it relates to the target keyword's actual search intent versus being a tangential topic the competitor happens to include, (b) roughly how much production effort it would take given our stated constraints — a paragraph addition versus a new section requiring original data or expert input versus something that needs a subject-matter expert we may not have. Do not treat every gap as equally worth closing; a subtopic only tangentially related to the core keyword is a lower priority even if it's easy to add, and a highly relevant subtopic that requires real effort is still worth flagging as a bigger investment, not skipped. STEP 3 — RECOMMEND Rank the gaps into a short "close these first" list of no more than five items, justified by the opportunity-versus-effort read, and a separate "consider later" list for the rest. WHAT NOT TO DO Do not recommend copying the competitor's specific framing, examples, or wording — describe what topic or angle is missing, never phrase it as a rewrite of their content. Do not inflate the gap list by treating minor phrasing differences as content gaps. OUTPUT FORMAT A coverage table (Subtopic | Our coverage | Relevance to keyword | Effort to close | Priority), followed by the ranked "close first" list and the "consider later" list as separate sections.
Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
A flat gap list is the default output shape here because listing missing subtopics is a simpler completion than jointly reasoning about relevance and production cost for each one, and GPT-5.1 will produce the simpler shape unless the effort dimension is made a required, per-row field rather than a closing suggestion — forcing it into the table structure itself is what prevents the model from mentioning effort only in passing for one or two items and dropping it for the rest. Distinguishing "partially covers" from genuinely covered matters mechanically because a keyword match between our page and the competitor's subtopic list would otherwise let a one-sentence mention on our page get silently credited as coverage, hiding a real content thinness problem that a manual audit would have caught — the instruction to treat shallow mentions as functional gaps closes that specific loophole. Scoring relevance to the target keyword's intent separately from raw subtopic overlap stops the model from recommending that every topic the competitor happens to cover gets added to our page; competitor content routinely includes topics that serve their own internal linking or monetization strategy rather than the shared keyword's actual intent, and without an explicit relevance filter those get inherited uncritically. The explicit ban on copying framing or wording addresses a real risk in this exact workflow: because the model is directly comparing two pieces of pasted content side by side, an unconstrained rewrite instruction would produce recommendations phrased so close to the competitor's actual sentences that acting on them verbatim would create a duplication or paraphrasing risk rather than genuinely original coverage.
What you get back
Subtopic: Tax deductibility rules | Our coverage: none | Relevance: high — directly tied to "how does a HELOC work" intent | Effort: medium, needs review-worthy accuracy but no licensed advisor required for general tax-rule explanation | Priority: close first.
Verified against
ChatGPT GPT-5.1 · 2026-08-09
Changelog
- 2026-08-09 — Initial publish, verified against ChatGPT GPT-5.1.
Building this for real?
This is a free starting point. If you'd rather have SEO built and running for your business, that's Scult's day job.
EXPLORE SEO
