Competitor Gap Report

The surface we map

A Competitor Gap Report maps the positioning gaps in what your named competitors have actually published over a recent window: what they over-claim, where their stated methodology is thin, which audiences they speak past, which formats they ignore, and the claims they make without evidence. The object is their public content, read against your ICP, to find the openings they have left.

Your own analytics show your traffic, not their positioning. A generic SEO tool reports keywords, not the argument a competitor is making and failing to make. This report reads the competitors' published content directly and names the specific gaps, which is the surface neither an on-platform dashboard nor a keyword tool puts in front of you.

The evidence standard

Every gap is anchored in one of exactly two evidence forms. Either a verbatim quote of up to 30 words from the competitor's crawled content, carrying the exact source URL, or an absence-of-evidence observation with a specific count, in the form "this competitor's crawled content over the window shows no mention of a specific topic across N published pieces." A competitor's claim is never paraphrased as if it were a quote, and no position, statistic, or methodology detail is invented.

This standard is mechanically enforced, not only requested. After the analysis is generated, an automated grounding gate checks every gap: each verbatim quote must be an exact substring of that competitor's crawled content and carry a source URL from their own pages, and each counted absence must match the number of pieces actually crawled in the window and name a topic that does not appear in their crawled content. Any gap whose evidence fails is dropped from the report rather than shown, and when fewer gaps clear the standard than usual, the report says so plainly instead of padding. What the gate verifies is the evidence, that each quote is real and each count is accurate. The reading of what a gap means for your positioning is our analysis, which is why every gap still arrives with its quote and source URL for you to check yourself.

How it works

1. Name the field. At intake you provide your company URL, your ICP, and three named competitors with their domains, plus an optional note on the competitive concern you want weighted.
2. Crawl the window. We crawl your homepage and each competitor's domain over a recent 30-day content window, recording each piece with its URL, date, content type, and an excerpt, and noting where a competitor's content is blocked or paywalled.
3. Identify the gaps. The analysis identifies the specific gaps it can ground, usually five to seven, across six categories (topics over-claim, methodology weakness, audience gap, format gap, positioning gap, evidence gap), each with a severity and a recommended response. It never pads the count: if only a few gaps clear the grounding standard, only those appear, with a note explaining why.

The category benchmark

Alongside your own gaps, each report places them against an anonymized aggregate of the prior Competitor Gap analyses we have run for companies in your category. The point is context: for each kind of gap, whether it is one the category broadly shares or one that is more particular to your company. The benchmark reports only how often a gap appears across those prior analyses. It never names another company and never ranks anyone.

Two lines are published. A gap is called category-common when more than half of the comparable analyses in your category show the same gap, and company-specific otherwise. Separately, a whole kind of gap is called category-open when at least two-thirds of the comparable analyses show a gap of that kind, which reads as the category broadly leaving that opening. Both figures are shares of prior analyses, nothing more.

The benchmark appears only once your category has enough prior analyses to stand behind a share, which we hold to a floor of at least eight. Below that floor, and whenever no category was selected at intake, the report says so plainly and shows no benchmark rather than an estimate drawn from too few analyses to be reliable.

• The aggregate counts analyses, not companies. A company we have analyzed in more than one quarter is counted once for each analysis, so the shares are weighted by analyses rather than by distinct companies.
• For a returning customer, the prior analyses in your category include your own earlier quarters. The aggregate is anonymized and cannot single out or set aside any one contributor, which is why we describe it as prior analyses in your category and never as other companies.
• The benchmark is a frequency within your category and nothing else. It is not a ranking of named competitors, not a company-versus-company comparison, and not a complete count of every company in your market.

Accuracy and limitations

• Version one is fixed to three named competitors. A wider competitive set is out of scope here.
• The window is the recent 30 days of published content. Older positioning is not captured, and a transient signal inside the window can be over-weighted.
• Coverage depends on crawl access. Where a competitor's content is blocked or paywalled, the report says so and notes that competitor may be underweighted, rather than guessing at what is behind the wall.
• An absence claim must cite a counted number of pieces reviewed. We do not write 'they have no good content'; we name the specific topic and the count of pieces it is missing from. That check runs against the content captured during the crawl, the lead portion of each page rather than a full-text search, so a topic mentioned only deep in a long page may read as absent.

What we do not claim

The gaps reflect what is publicly published in the window, not a competitor's private roadmap or internal strategy. We do not assert what a competitor will do next, and we do not guarantee that exploiting a gap will win the position. The report names the opening and shows the evidence; the strategy and the execution are yours.