An influencer brand suitability review should compare dated public evidence with written campaign constraints. Define those constraints before reviewing candidates, separate observed conduct from interpretation, and send uncertain cases to a named human reviewer. Personal dislike, a creator's identity, and an unsupported allegation should never become interchangeable reasons to reject someone.
Define what this campaign cannot carry
Start with the product, intended viewers, planned content and paid distribution. Then write the conduct or content that would conflict with that campaign. "Controversial" is too vague to apply consistently. "The sponsored tutorial must not encourage unsafe use of the product" gives reviewers something they can assess.
Separate three questions in the review record:
- Does the proposed content meet the relevant platform rules?
- Does the evidence meet a campaign-specific exclusion rule?
- Does the reviewer personally dislike the creator's style or opinion?
Only the first two belong in a suitability decision. A creative preference can remain a separate selection criterion when it relates to the brief. For example, a spoken product demonstration may require clear explanations. That does not make a creator who prefers silent films a reputational risk.
YouTube's advertiser-friendly guidelines illustrate why scope matters. They concern ad eligibility and cover the content, title, thumbnail, description and tags. They also distinguish context, including news reporting and educational treatment of violence. A keyword appearing in a video cannot settle the review by itself. These platform rules do not certify that a creator suits your campaign.
Complete the policy before opening profiles
The worksheet below is a hypothetical policy for a home-tool tutorial campaign. Its review window and deadlines are operating choices, not platform requirements or legal standards. Adjust them to the campaign and apply the same version to each candidate.
| Policy field | Hypothetical completed entry |
|---|---|
| Campaign | Home-tool tutorial showing use according to the manufacturer's instructions |
| Content exclusion | Sponsored content must not demonstrate or encourage bypassing safety devices |
| Creator-history rule | Reject for this campaign if a verified original post within the review window encourages that conduct and the creator confirms it remains their intended approach |
| Review window | Public posts from the preceding six months, plus specific older evidence referred for review |
| Evidence required | Original public URL, publication date, review date, relevant timestamp and a context note |
| Out of scope | Religion, ethnicity, disability, sexual orientation, inferred political identity, and unrelated personal taste |
| Uncertainty route | Campaign lead reviews ambiguous context; privacy or legal questions go to the qualified owner |
| Decision deadline | Resolve before the partnership offer; no unresolved case receives approval |
| Recheck | Review new relevant evidence and the planned sponsored content before launch |
| Retention | Privacy owner sets access, retention period and deletion date for the record |
An older post entering review needs a reason, such as a specific report tied to the same conduct rule. Avoid expanding the search window solely because one reviewer dislikes a candidate.
For the mechanics of selecting a consistent post sample, use a fixed recent-post review. Keep that sampling decision separate from the suitability rules.
Write evidence that another person can inspect
Each concern needs a short record with four parts:
- Observed evidence. Identify the original post, its date and the part you reviewed. Describe what the creator said or did.
- Context. State whether they were demonstrating, reporting, quoting, criticizing or joking. Read or watch the surrounding material.
- Campaign connection. Name the written rule and explain the possible conflict.
- Open question. State what remains unknown and who can resolve it.
Replace "reckless creator" with a statement such as "the video shows removal of a safety guard, but the excerpt does not establish whether the creator endorses it." That hypothetical note identifies both an observation and a limit. It leaves the reviewer a task.
A reposted screenshot or a forum allegation can prompt a check. Treat it as unverified until you establish attribution and context. If the original is unavailable, record that limit. A deleted post does not establish why it disappeared.
Do not put follower counts, engagement estimates or audience demographics into a moral-risk score. This worksheet uses content evidence. It makes no inference about a creator's character from their audience.
Keep protected traits out of the suitability score
Exclude protected traits and sensitive personal inferences from this operational review. Do not guess religion from clothing, ethnicity from a surname, disability from appearance, or political identity from a list of followed accounts. Do not ask a model to produce those guesses for you.
There is a specific privacy reason for caution in the UK. The ICO's special category data guidance includes ethnicity, religion, political opinions, health and sexual orientation. It explains that intentionally inferring such information, or treating someone differently on that inferred basis, can engage special category requirements. Confidence in the guess does not remove that issue.
The ICO page is under review following the Data (Use and Access) Act. Its guidance is UK-specific. Ask the responsible privacy or legal adviser about the applicable rules before collecting sensitive data. This worksheet is an operational recommendation, not legal advice.
Reviewers can document relevant conduct without assigning an identity label. A specific public threat can warrant escalation under a written conduct rule. A creator discussing their own religion does not, by itself, establish a campaign conflict. Audience hostility toward a creator's identity should not become evidence against that creator.
Record accept, reject and escalate decisions
All cases below are hypothetical. They apply the home-tool policy above and describe no real creator.
| Decision | Evidence and rule | Documented outcome |
|---|---|---|
| Accept | A reviewer dislikes the creator's criticism of an unrelated film. The post has no connection to the campaign rules. | Accept for suitability. Record that no campaign exclusion applies; assess creative fit separately. |
| Reject | A dated original tutorial within the window encourages bypassing a safety device. The creator confirms that approach would remain in the proposed tutorial. | Reject for this campaign under the written safety rule. Record the source, confirmation and decision owner. |
| Escalate | A circulating excerpt shows a removed safety guard. The complete video and the creator's intended lesson are unclear. | Hold the decision. Ask for the original context and a description of the proposed demonstration. |
The rejection belongs to this campaign and policy version. Avoid turning it into a permanent company-wide label. An exclusion list with review dates can help manage decisions that need to carry into later searches.
Give uncertain cases an owner and an answer
Assign someone other than the first reviewer to disputed cases. Give them the evidence, the rule, the uncertainty and the deadline. If the team cannot resolve the evidence before the deadline, record "decision deferred, evidence unresolved" rather than a confirmed violation.
Ask the creator a narrow, factual question. In the hypothetical safety case: "Could you share the full video and explain whether the sequence demonstrates a practice you recommend? Our campaign requires following the manufacturer's safety instructions."
Modash's discussion of influencer risk recommends talking through concerns with creators, including missing context in earlier posts. A reply adds evidence; it does not automatically clear or condemn the creator. Save the relevant clarification and explain how it changes the decision.
Before reviewing the next candidate, have two reviewers apply your worksheet to the same hypothetical case independently. If their decisions differ, rewrite the rule or evidence requirement before using it on a real creator.
