Put an exclusion in your search brief only when a confirmed failure would make the planned collaboration unworkable. Keep aesthetic preferences, broad category labels and uncertain audience signals in a separate review list. Treat missing evidence as unresolved. Before using negative keywords, inspect the candidates those words would remove.
That is the core test for influencer exclusion criteria. A search brief should explain what rules out a candidate, what merely lowers their priority, and what needs confirmation before a decision.
Give each rule a consequence and an evidence standard
Ask two questions before making a rule mandatory:
- What specific campaign requirement fails if we accept this creator?
- What evidence would establish that failure?
If the answer is only "we prefer a different look," record a preference. If the creator confirms they cannot attend the required event, record a hard failure for that campaign.
Modash's article on influencer search mistakes warns against making every fit preference non-negotiable. The operational response is to give each rule a different job. Hard rules remove confirmed mismatches. Soft rules help order the remaining candidates. Unknowns create a follow-up task.
Use this hypothetical brief for a reusable lunch-container launch:
| Criterion | Classification | Evidence and action |
|---|---|---|
| Must film the supplied product before the fixed launch date | Hard constraint | Ask the creator to confirm availability. A confirmed inability rules them out for this launch. |
| Must demonstrate packing and cleaning the container | Hard constraint | Confirm willingness and ability to perform the required demonstration. |
| Prefer food-focused accounts | Soft preference | Review relevant posts. Keep adjacent creators who demonstrate the same use case. |
| Prefer a neutral kitchen setting | Soft preference | Record the production preference without excluding other settings. |
| No current competing commitment that prevents this collaboration | Hard constraint, pending confirmation | Ask about relevant commitments. A competitor mention alone does not establish a conflict. |
| Audience location evidence has not arrived | Unresolved | Request the relevant report. Do not enter zero or mark the creator unsuitable. |
A hard constraint can remain unresolved during discovery. It becomes an exclusion only after evidence confirms a failure. If a required check remains unresolved at the booking deadline, record "requirement unverified by deadline" separately from "confirmed mismatch."
Keep campaign exclusions scoped to the campaign. Unavailability this month should not become a permanent rejection. When you need to investigate a possible competing relationship, use the sponsorship-history review process before assigning a reason.
Write rules against the field you mean
"Wrong location" is too vague to be an exclusion. It could mean the creator cannot reach an event, the product cannot reach their address, or their audience is outside the target market. Those require different evidence.
Audience fields also have different meanings. YouTube's audience documentation describes top geographies in terms of channel watch time. TikTok Studio's documentation separates viewer analytics from follower analytics, with demographic information in both sections. A rule about followers cannot silently use viewer or watch-time data as its substitute.
For any audience threshold in a brief, write down:
- The population being measured, such as followers or viewers.
- The metric and reporting period.
- The source and collection date.
- What happens when that field is unavailable.
These reports are creator-side analytics. Their documentation does not establish that every external researcher can access another creator's reports.
Keep three evidence states distinct. An observed post can support a statement about that post. An inferred audience attribute needs the provider's methodology and stated limits. An unavailable field remains unknown. This guide uses no third-party audience estimates.
YouTube explains that privacy and system limits can produce missing data, dashes or zeros. Low traffic or a selected dimension can also restrict a report. A blank demographic field therefore cannot establish that the audience lacks the required group. For the follow-up decision, use the guide to evaluating creators with missing audience data.
Test what each excluded word would hide
A keyword is a clue about language. A rejection rule needs a reason about the collaboration. The gap becomes visible when you inspect the text around a match.
The following examples are hypothetical. They show mistakes a literal exclusion could cause; they do not describe any particular search product's matching behavior.
| Excluded term | Candidate it could hide | Better instruction |
|---|---|---|
| "travel" | A commuter who posts about packing lunch for train travel | Review whether the creator can demonstrate everyday meal transport. |
| "kids" | A parent who packs their own work lunch alongside children's lunches | Check the intended user and demonstration in relevant posts. |
| "plastic" | A creator discussing alternatives to disposable plastic | Read the sentence and product context before deciding. |
| A competitor's name | A creator making an unpaid comparison | Review the post, then confirm any commitment that affects this project. |
| "luxury" | A creator describing a low-cost alternative to a luxury purchase | Judge the demonstrated price context and audience need. |
Before activating an exclusion, establish where the search looks. Does it match bios, captions, transcripts, account names or category labels? Does it match phrases, individual words or word fragments? Check the tool's documentation or inspect its returned matches. If you cannot establish the behavior, use the term as a manual review flag.
For new positive terms that describe the use case, work through turning a customer problem into creator search terms. That keeps the exclusion list from becoming the main definition of whom you want.
Audit one exclusion at a time
Keep a baseline candidate list before applying the disputed exclusion. Then compare the baseline with the list after that single rule. Hold the other criteria constant and record when you ran the comparison.
Review the removed candidates in two groups: clear failures and possible false exclusions. For each possible false exclusion, inspect the matched field and decide whether the actual campaign requirement fails. Record unresolved cases separately.
For example, suppose a hypothetical baseline contains 60 creators. Excluding "travel" removes 12. Reviewing those 12 finds four commuter-lunch creators who still meet the brief, six unrelated vacation accounts, and two profiles needing more evidence.
That review found four false exclusions among the 12 removed candidates. It does not show that the rule will make the same proportion of errors across all searches. The appropriate change is to remove the broad keyword ban or replace it with a narrower, reviewable condition.
This audit is a recommended workflow, not a platform rule. Repeat it when the brief changes or a recurring rejection reason looks broader than its evidence.
Keep a rule register beside the brief
For each proposed exclusion, complete this record:
| Field | What to enter |
|---|---|
| Requirement | The task or campaign condition that must hold |
| Rule type | Hard constraint, preference or unresolved check |
| Evidence needed | The specific confirmation, report or post review |
| Search behavior | Field searched and how the term matches |
| Missing evidence | Follow-up owner and decision deadline |
| Exclusion scope | This campaign, date window or other defined scope |
| Audit result | Confirmed failures, false exclusions and unresolved cases reviewed |
| Review date | When the rule should be reconsidered |
Start with the broadest negative keyword in your current brief. Review the candidates it removes, then replace the keyword with the campaign requirement you meant to enforce.
