Competitive intelligence for ecommerce and digital teams
See how your digital presence compares with your competitors
CompComp (Compare Competitors) analyses the exact part of your web presence you choose - a host, a subdomain, a category path or a single page - and compares it with competitors using the same criteria. Every conclusion is marked observed, inferred or unknown and links back to the evidence behind it.
- The same criteria for everyone
- Your own site and every competitor are assessed against the identical rubric, so differences come from the sites, not from who did the analysis.
- The exact scope you choose
- A host, a subdomain, a category path or a single page. shop.com/product stays that path and its children - never all of shop.com, and never /products.
- Evidence before completeness
- Every conclusion is marked observed, inferred or unknown and links to the evidence behind it. Unknown is a valid result and is never counted as a gap.
How CompComp works
Step 1
Choose what to compare
Define your own analysis target as a precise URL scope rather than a brand name or a bare domain.
Step 2
Add your competitors
Add competitor targets manually, or use Find similar competitors - suggestions come with a reason, evidence and a confidence level, and unresolved domains stay unresolved.
Step 3
Analyse
CompComp collects site evidence inside the scope, then runs each analysis area step by step so you can follow progress and see what failed if something does.
Step 4
Compare in the Matrix
Companies as columns, analysis areas as rows, with Strong, Moderate, Weak or Unknown for a quick read and detail one click away.
Step 5
Verify in Evidence
Every assessment links to the pages and excerpts it was based on, with retrieval timestamps, so a conclusion can be checked rather than trusted.
Inside the product
See what CompComp delivers
See how CompComp moves from defining competitors to side-by-side comparison and detailed evidence-backed analysis.
Every participant, every area, one view
Companies as columns, analysis areas as rows. Your baseline stays highlighted, coverage is shown next to each assessment, and Unknown is reported as missing evidence rather than a gap.
Screenshots show the CompComp interface using a demonstration comparison. Company names, domains and findings are fictional.
Want to see this applied to a real scenario? Read the brand channel comparison use case.
What CompComp analyses
Each area answers a business question, and each is assessed the same way for your own site and for every competitor in the comparison.
Who does this site appear to be written for?
Audience
Audience and ICP signals inferred from the analysed pages. Always labelled as inferred - a public site never proves an internal ICP.
What does this company say it does differently?
Positioning
How clearly the offer, the audience and the differentiation are stated inside the scope, and how consistently they are repeated.
Can this scope be found and understood in search?
Search readiness
Indexability, on-page structure, metadata and structured data inside the exact scope you selected.
Is this content prepared to be used by AI answers?
AI Visibility signals
A readiness assessment based on what is on the page. CompComp does not query ChatGPT, Gemini or Perplexity and never claims a ranking or a citation.
Is the content structured for generative discovery?
GEO readiness
Machine-readable structure, answerable content, attribution signals and crawler access for generative engines.
Could an agent buy from this site on the customer's behalf?
Agentic Commerce readiness
Product identity, variants, attributes, price and offers, availability, merchant details, delivery, returns, reviews and purchase-path clarity. Product schema alone is not enough.
What does the comparison add up to?
SWOT
Strengths, weaknesses, opportunities and threats drawn from the evidence already collected, not written freehand.
Built for comparison, not just auditing
An audit asks
"How well does this one site score against a checklist?" The answer is useful for a backlog, but it says nothing about the companies you are actually competing with.
CompComp asks
"How does this scope compare with these specific competitors, on the same criteria, on the same day, with the evidence attached?"
- The same criteria for everyone. Your own site and every competitor are assessed against the identical rubric, so differences come from the sites, not from who did the analysis.
- The exact scope you choose. A host, a subdomain, a category path or a single page. shop.com/product stays that path and its children - never all of shop.com, and never /products.
- Evidence before completeness. Every conclusion is marked observed, inferred or unknown and links to the evidence behind it. Unknown is a valid result and is never counted as a gap.
- No invented measurements. AI visibility and GEO are readiness assessments based on what is on the page. No answer engine is queried, and no ranking or citation is claimed.
Where CompComp sits
How CompComp's comparison-first approach differs from point solutions that audit one site at a time.
An illustrative example
Illustrative only - not a real customer result
A retailer sets their baseline to their running-shoe category path and adds the equivalent category paths of three competitors. Nothing outside those paths becomes content evidence.
- Search readiness would show which category pages are structured and described well enough to be understood.
- Positioning would show whether each retailer states who the range is for, or only lists products.
- Agentic commerce readiness would show whether price, availability, delivery and returns are machine-readable.
- Unknown results would stay unknown - for example where a competitor blocks automated access.
Who CompComp is for
- Ecommerce teams
- Understand where a category or product experience is genuinely behind, and where it only feels behind.
- Brand and marketing
- Check whether the positioning you believe you own is actually visible on the pages people and machines read.
- Digital leaders
- Bring a repeatable, checkable comparison into planning discussions instead of assorted screenshots.
- Consultants and agencies
- Apply one method to every client and competitor, and show the evidence behind each conclusion.
Built to make comparison more defensible
- Every finding is tied to evidence you can open.
- Observed, inferred and unknown are separated, not blended.
- Unknown means the check could not be completed - it is never presented as a competitive gap.
- The analysed scope is stated explicitly, and materially different scopes raise a warning.
- Known limitations are written down rather than hidden.
Read why CompComp exists, or get in touch with a question about the method.