Audience research
Part of Marketing attribution audience research: six populations
Researching attribution competitors without creating false history
Research marketing attribution competitors through dated public evidence, source classification, careful claims analysis and explicit unknowns.
Marketing attribution competitor research can establish what a vendor publicly claims, which legal entity is visible and how an offer is described on a given date. It cannot reveal customers, revenue, market share, product performance or popularity unless a suitable independent record measures that exact fact.
What to take away
- Competitor research can show public claims and legal entities, but not customers, revenue or market share.
- Define the buyer question, geography, cut-off date and qualifying service before searching.
- Classify records as first-party claims, legislation, independent tests or editorial interpretation.
- Personal data duties apply even when officer or employee details are publicly visible.
- Stop if legal identity is unclear or unsupported figures would be required.
Define the comparison before searching
- State the buyer question, the target geography, the research cut-off date and the qualifying attribution service.
- Include a manual or existing-process baseline so the work does not assume a purchase.
- Fix common fieldslegal party, exact product or service, method, supported data scope, evidence output, buyer work, contract unit, privacy, security, accessibility, support, export, deletion and exit.
- Exclude unrelated analytics products, duplicate brands and records whose edition or date cannot be identified.
Set the research cut-off date before searching, record it in every output, and apply it to each record. This is a desk method for research into marketing attribution and analytics vendors, not a ranking: no competitor account was opened and no service was tested.
Start from the attribution vendor set
Competitors here are the platforms a buyer would shortlist. Web and app analytics include Google Analytics 4 and Adobe Analytics; mobile measurement partners include AppsFlyer, Adjust, Branch, Singular and Kochava; product analytics include Mixpanel and Amplitude; and multi-touch attribution and incrementality tools include Northbeam, Triple Whale, Rockerbox and Measured.
- Name the vendor set before searching and record why each platform is in or out.
- Note the segment each vendor servese-commerce, app, lead generation or enterprise marketing.
- Record how each vendor sells in the United Kingdomdirect, through agencies or through resellers.
Define the comparison first
- Buyer question, England rule, cut-off date
- Manual baseline before assuming a purchase
- Fix common fields across every record
- Exclude unrelated analytics and duplicate brands
- Evidence cut-off6 September 2026
- Desk method, not a provider review
Classify every record
- Mark supplier pages, terms and documentation as first-party claims.
- For each attribution vendor, treat the product documentation, pricing page, changelog, status page and integration list as separate first-party records with their own dates.
- Label legislation, regulator guidance, industry frameworks such as the IAB Europe Transparency and Consent Framework, and public registers separately.
- Record independent tests only when their method, date, product version and population are available.
- Put editorial interpretation in its own field and record unknown when evidence is absent rather than filling gaps.
Companies House's free company-information service can provide items such as registered address, incorporation date, officers and filed documents. Those register fields can help identify the legal entity behind an offer, but they are not independent evidence of current trading activity, product quality or availability of the attribution product in the target market.
Classify each record by source
Source type
- Supplier pages
- First-party claims only
- Terms and docs
- First-party claims only
- Legislation, guidance
- Separate labelled source
- Public registers
- Legal entity, not quality
- Independent tests
- Only with method and date
- Editorial view
- Own field, not evidence
What it proves
- Supplier pages
- Terms and docs
- Legislation, guidance
- Public registers
- Independent tests
- Editorial view
Capture pages without creating false history
- Save the visible URL, page title, access timestamp, product edition, territory and relevant excerpt.
- Keep a dated public copy, using the Internet Archive's Wayback Machine or archive.today, and cite the archived URL next to the live one.
- Record redirects, login barriers and missing dates.
- Revisit material claims before publication rather than relying on an old screenshot.
- Respect site terms, technical restrictions and licence conditions. Do not bypass access controls.
Copying a public page is not automatically unrestricted. The UK government's current intellectual-property management guide explains that database rights are distinct from copyright and may protect database contents where the statutory conditions are met. In the UK those database rights come from the Copyright and Rights in Databases Regulations 1997, which sit alongside the Copyright, Designs and Patents Act 1988. Qualified IP advice is required for the proposed collection and reuse.
Treat personal data as personal data
- Avoid collecting officer or employee details unless the declared purpose needs them.
- Record purpose, lawful basis, transparency approach, access, retention and deletion for any personal data.
- Do not infer protected characteristics, influence or buying authority from names and profiles.
- Keep research contacts out of later sales lists unless that separate use has its own lawful route.
The ICO's lawfulness, fairness and transparency page says organisations need a valid basis and must use personal data fairly and openly. Public visibility does not by itself answer those duties. For anything that tracks users across a site or app, the ICO also enforces the Privacy and Electronic Communications Regulations (PECR), which require consent for storing or reading information on a device.
Test claims, then report the gap
- Preserve the supplier's exact wording and likely audience impression.
- Check attribution claimsthe model used, modelled or blended conversion figures, incrementality tests, marketing mix modelling, data retention and integrations — against the method, date and population the vendor publishes.
- Seek evidence matching the claimed product, population, conditions and date.
- Do not turn case studies or testimonials into a typical result.
- Withdraw any comparison that cannot be kept like for like.
The ASA's substantiation guidance says objective advertising claims capable of substantiation need adequate evidence before publication. A supplier claim is evidence that the claim was made, not proof that it is true. That applies to accuracy, incrementality and return-on-investment claims in attribution marketing.
Finish with a coverage table listing included and excluded suppliers, unknown fields, source types, conflicts and the person responsible for each verification. Stop if legal identity is unclear, the scope cannot be normalised, collection would breach an access condition, or unsupported figures would be required to produce a result.
Before you act
- State the buyer question, geography and research cut-off date.
- Include a manual baseline so the work does not assume a purchase.
- Save the visible URL, access timestamp and product edition.
- Avoid collecting officer or employee details unless needed.
- Preserve the supplier's exact wording and likely audience impression.
- Withdraw any comparison that cannot be kept like for like.
Common questions
What can competitor research establish about a supplier?
It can establish what a supplier publicly claims, which legal entity is visible and how an offer is described on a given date. It cannot reveal customers, revenue, market share, product performance or popularity unless a suitable independent record measures that exact fact.
How should personal data be handled in this research?
Avoid collecting officer or employee details unless the declared purpose needs them. Record purpose, lawful basis, transparency approach, access, retention and deletion for any personal data. Do not infer protected characteristics, influence or buying authority from names and profiles. Keep research contacts out of later sales lists unless that separate use has its own lawful route.
When should the research stop?
Stop if legal identity is unclear, the scope cannot be normalised, collection would breach an access condition, or unsupported figures would be required to produce a result. Also withdraw any comparison that cannot be kept like for like.
