A content gap analysis prompt can produce a credible gap list from evidence you paste yourself: your competitors' page titles, their headings, your own URL list, and the questions your customers ask. What it cannot produce is traffic, rankings, or authority. Those need a paid tool, and pretending otherwise wastes an afternoon.
So the honest version is narrower. Use prompts to read competitors properly, find the questions nobody answers, and order your build list. Use a tool for volume and difficulty, or skip them and judge from your own sales conversations.
What you can and cannot learn without a subscription
You can read every page a competitor publishes. You can map their headings, note which topics they have covered twice, and see where their articles stop. You can list the questions their content never answers. You can compare that against your own coverage and your own customer questions. That is enough to build a defensible plan.
You cannot see their traffic, which keywords earn them clicks, how many links point at a page, or how their pages perform in search. Any prompt that claims to produce those will either invent them or quote a figure it cannot trace. Both are worse than nothing, because they look like evidence.
One workaround is genuinely available. Ask several assistants the same buyer questions and record which sites they cite. That is a manual, imperfect sample of what the answer layer already favours, and it costs an afternoon rather than a subscription.
The content gap analysis prompt that works from pasted evidence
Act as a content strategist for [business type] serving [audience and area or market]. Here is my current content: [paste URL and one-line purpose for each, or write "almost none"]. Here are my competitors' pages: [paste page titles and URLs, grouped by competitor]. Here are the questions my customers actually ask: [paste from sales calls, support email, or your own notes]. Produce three outputs. First, a coverage table: topic, which competitors cover it, whether I cover it, whether my customers ask about it, and the gap type. Gap type must be one of: unanswered, covered poorly by everyone, covered well by me already, or not worth building, with a one-line reason. Second, a list of questions from my customer list that no competitor page appears to answer. Third, a ranked build order with the reason each item is above the one below it. Rules. Only use the pages and questions I supplied. Never estimate traffic, keyword difficulty, domain authority, or how well a page ranks. If you cannot tell whether a page covers a topic from its title, say "unknown" rather than assuming. Do not recommend building a page I already have. Do not recommend a page I have no expertise to write.
Fill in these details
- Business type and area: Local businesses compete on geography, and a national list will mislead you.
- Current content: Paste real URLs. "Almost none" is a valid and common answer.
- Competitor pages: Titles and URLs are enough for a first pass. Headings make the second pass sharper.
- Customer questions: The highest-value input. Real language beats any term list.
- Your expertise: The prompt cannot know what you can credibly write about.
Spend thirty minutes on the input. Most disappointing analyses come from thin evidence, not from a weak prompt.
A second prompt that turns the gap list into a build order
Take this gap list and produce a build plan for [number of weeks] weeks at [pieces per week] pieces. For each item, give: the question it answers, the page type, the sections it must contain, what I would need to research before writing it, and whether it can be written from what I already know. Put the items answerable from existing knowledge first, because they ship fastest and give the site structure to link to. Then mark every item that depends on a fact I do not have, and name the specific source type that would resolve it. Do not fill gaps with plausible content. Do not add items that were not in the list. List: [paste].
This ordering rule matters more than it looks. Sites built on pages you can write immediately outrun sites that wait for every expert interview to land.
Worked example: a local service business
A bookkeeping firm in a mid-sized city lists eight own pages and eleven competitor pages. The evidence-only analysis produces this:
| Question | Competitor coverage | My coverage | Gap type |
|---|---|---|---|
| What records do I need to close a month | Two thin posts | None | Unanswered, worth building |
| My client asked for a receipt, what now | Not covered | One short post | Covered poorly by everyone |
| How much does a bookkeeper cost locally | One post, generic | None | Build, but needs real local pricing |
| Which bookkeeping software integrates with my bank | Not covered | None | Unanswered, needs research |
The fourth item is flagged for research because the answer depends on bank partnerships the firm cannot verify. Building it from guesswork would produce a page that ages badly. The second is a quick improvement to a page that already exists, which is a better first move than a new URL.
How to check a gap analysis before you build anything
1. Open every competitor page listed
Titles lie and pages get merged. If the model inferred coverage from a title, confirm it on the page. Where it cannot tell, it should say unknown, and you should treat unknown as uncovered until you check.
2. Confirm the gap is a gap for your reader
A topic no one in your market asks about is not an opportunity. Check the gap against your customer questions and against what your sales team already explains on repeat.
3. Check you can write it
You need the expertise, the access, or the examples. A page written from general knowledge competes against pages written from experience, and it will read like it.
4. Delete what you already cover
The most common failure is building a second page on a topic you have already answered. Search your own site before accepting any row marked as a gap.
Common content gap analysis mistakes
- Accepting traffic or difficulty figures from a prompt: They are invented. Always.
- Copying a competitor's angle: Their page wins for their reasons, which you cannot see. Cover your own question better.
- Ignoring topics you already own: Updating an existing page is usually cheaper and faster than building.
- Building on unverified facts: Local prices, legal steps, and integrations change. Verify before writing.
- Treating a big list as a strategy: Twenty gaps is not a plan. Three pages you ship beat twenty you queue.
- Skipping the customer questions step: Without it you are writing for competitors rather than for buyers.
Frequently asked questions
Can content gap analysis be done without Ahrefs or Semrush?
For deciding what to write, yes. Reading competitor coverage and matching it against your own content and your customers' questions does not require a subscription. Judging whether a topic has enough demand to justify the effort usually does.
What is the fastest way to collect competitor pages?
Search your main topic, then the three or four questions beneath it, and record every result that is not you. Twenty minutes of searching usually produces a usable evidence base.
Is copying a competitor's content a valid gap strategy?
No, and it is rarely even effective. What you can borrow is the question they answered, not the answer. Write it from your own evidence.
How often should I rerun the analysis?
Once a quarter, and after any large change to your service or market. More often than that, the output will be noise rather than a signal.
Run one gap analysis, then pick three pages
Collect the evidence, run the prompt, and take only the top three items you can write from what you already know. Ship those, then rerun with what changed. For more prompts covering the search work around your content, browse the SEO optimization prompt collection.