Food influencers' partnership review methods

What to take away

Enforcement notices and official statistics are now published online, so a review can point to a checkable source rather than a vendor's own wording.
Vendor material describes a product; it does not test it. Reviews here are documentary, not hands-on.
A method that scores disclosure, audience fit and measurement first will rank platforms differently from one that scores price first.
Each score must name its source, so a later review can be rechecked against a dated document.
Left out on purposetaste tests, site visits, mystery shopping and creator interviews.

What gets assessed

A methodology for reviewing food influencers' partnership platforms sets the criteria before anyone opens a media kit. Every claim in a later review traces to one of three inputs: the vendor's published material, an independent source, or an editorial judgement made against a written rule. Record the source's title, publisher and publication or update date; if a source cannot be identified or dated, mark the claim unverified.

Three inputs for every claim

Vendor's published material
Independent source
Editorial judgement against written rule

That separation matters because platforms describe themselves in the language of reach and engagement. Those figures are self-reported until a third party checks them. Which? publishes consumer data resources that can add context about wider consumer behaviour, but they cannot verify a platform's own audience or engagement figures.

The criteria are audience relevance, disclosure practice, measurement quality and contract terms, and total cost.

What gets left out

This method does not include taste tests, site visits, mystery shopping or interviews with creators. None was done, and claiming otherwise would break the line between evidence and assertion. A platform's demo video is not a trial. A case study supplied by the vendor is not independent evidence.

What is included is documentary: published terms, regulator guidance, official statistics relevant to the claim and the platform's own help pages. Name each document, its issuing body and publication or update date; where no suitable independent source exists, record the gap. Where a review scores something, the score reflects the strength of the supporting source. The reviews and comparison methods for food creator partnerships guide explains how the criteria were built and why some carry more weight.

Claims against evidence

Each vendor claim sits next to the source that supports or contradicts it. Check claims about personal-data handling against the UK GDPR and Data Protection Act 2018; where electronic marketing or cookies are involved, check PECR too. Identify any relevant ICO guidance or decision by title and date. The ICO's enforcement action concerning TikTok's handling of children's personal data is an example of regulator action, not evidence about a different platform's conduct.

For disclosure claims, the method names the regulators before it scores anything. The CMA's note Who regulates hidden advertising lists three bodies with responsibility for making sure paid-for endorsements are properly labelled on social media: the Advertising Standards Authority, Trading Standards Services and the Competition and Markets Authority. It explains that from 6 April 2025, Part 4 of the Digital Markets, Competition and Consumers (DMCC) Act replaced the Consumer Protection from Unfair Trading Regulations 2008, and that the CMA and Trading Standards enforce those Part 4 rules against content creators, brands and intermediaries such as agencies. On the advertising-code side, rule 2.1 of the CAP Code says marketing communications must be obviously identifiable as such. The CMA's guidance for brands adds the test a reviewer can apply to a creator's posts: a proper advertising label should be clear, upfront, prominent and timely, and it recommends 'Ad' or 'Advert'.

A claim about staff injury patterns may be contextualised with HSE statistics, but national data cannot verify a particular restaurant's hygiene, safety or insurance. Require evidence that directly supports a claim about an individual restaurant or partnership; otherwise, mark it unverified. Where no independent source exists, the review says so. A gap is recorded as a gap, not filled with an assumption.

Scoring and weighting

Scores run across five areas: audience fit, disclosure practice, measurement quality, contract terms and total cost. Define the scale before assessment: label a claim supported when a named, dated source directly supports it, mixed when evidence is incomplete or conflicting, and unverified when evidence is absent or vendor-only. Show the result for each area and disclose any weighting; if the scoring rules or weights are hidden, do not trust the overall ranking. Audience fit and disclosure are gates: if either cannot be verified, do not call the partnership recommended, whatever its cost or dashboard score.

This method's own weighting is published so it can be checked. All five areas are scored on the supported, mixed and unverified scale. Audience fit and disclosure are scored first and also act as gates; measurement quality and contract terms carry equal weight with each other; total cost is scored last and cannot lift a partnership that fails a gate. A disclosure gate passes only when the creator's past paid posts carry a label that is clear and upfront, such as 'Ad', rather than a tag, discount code or vague hashtag.

Cost is scored last on purpose. A cheap partnership that breaches advertising rules is not cheap. For example, if a hypothetical restaurant campaign relies on a platform media kit for its local-audience claim and has no independent corroboration, mark that claim unverified rather than treating high claimed reach as a pass. The product comparison for food creator partnerships applies comparison criteria to products; this methodology article does not assess a named platform, creator or restaurant partnership.

Due diligence checklist

Confirm the creator or platform publishes a clear paid-partnership disclosure. Use the CAP Code as the reference and check relevant ASA guidance or rulings; the ASA is the UK's self-regulatory organisation for advertising.
Check personal-data processing against the UK GDPR and Data Protection Act 2018; where electronic marketing or cookies are involved, check PECR too. Record the applicable lawful basis.
Ask for the platform's complaint or takedown process, not a verbal promise.
Verify any safety, hygiene or insurance claim against a published source that directly supports it before repeating it.
Separate self-reported reach from any figure a third party has audited.
Note the contract length, exit terms and who owns the content after the campaign ends.
Keep a dated copy of every source used, including its title and issuing body, so a later review can be rechecked.

Common questions

What was assessed?

Documentary evidence only: vendor materials, regulator guidance, published statistics and platform documentation. No product was trialled and no creator was interviewed.

How are vendor claims treated?

They are recorded as claims, not findings. Each one is matched to an independent source where one exists, and flagged as unverified where none does.

Why does disclosure carry so much weight?

A missing disclosure is a regulatory and reputational risk that no engagement figure offsets. It is scored heavily for that reason.

Can a score change after publication?

Yes. If a platform changes its terms or a regulator updates guidance, the review is rechecked and the score revised with the new source dated.