Give the agent the operational context
State the product, program, target creator, platform, region, evidence requirements, and exclusions.
Transparent screening
Score creator fit with reasons your team can inspect.
UGC Agent turns profile evidence into an explainable first-pass score. The score helps prioritize manual review; it does not predict creative performance, campaign results, or whether someone will accept a project.
Explainable scoring reasonsFollower and language guardrailsCampaign-specific relevance
Workflow
From request to usable output.
Inspect the structured output
Review sources, score reasons, classifications, risk flags, and any missing evidence before acting.
Keep the human decision
Use automation for the first pass while your team controls outreach, approvals, commercial terms, and publishing.
Explainable scoring
A score your team can actually inspect.
Every retained creator keeps the reasons behind the number, not just the number.
13/15Fit score
- UGC positioning in handle and bio
- Public email listed, personal inbox
- Portfolio link resolves to a demo reel
- Niche and follower range match the brief
Illustrative score card for a placeholder creator, styled after the real scoring reasons attached to each lead.
01
What the baseline score measures
The baseline rubric awards signals for UGC language in the username, display name, or bio; public contact information; portfolio language; known portfolio-link hosts; and structured niche descriptions. Each retained lead carries the reasons behind its score.
02
What can exclude a profile
Default micro-UGC sourcing excludes private accounts, profiles outside the configured follower range, strongly non-English bios, missing emails, low scores, and profiles that fail the campaign relevance check. Brand-collaboration mode uses a wider follower band and requires public collaboration evidence.
03
Why scoring is not a verdict
A profile can look qualified and still be wrong for a particular product, rate, timeline, or creative direction. Use the score to reduce the first-pass workload, then review content examples, brand safety, availability, compensation, and rights manually.
04
Adapt the score to the job
Saved agent preferences, one-run overrides, seed-creator analysis, niche requirements, agency-email exclusions, and brand-evidence mode change which profiles survive. The goal is consistent screening around the campaign, not a universal ranking of creators.
05
How relevance changes the baseline score
Baseline UGC signals show whether a public profile presents as a contactable creator, but campaign relevance answers a different question: does this person fit the product, buyer, region, format, and creative context in the current brief? A strong generic profile can still fail a specific relevance gate.
Seed analysis and saved preferences can add context without rewriting the scoring system into a hidden universal ranking. The result should explain which requirements were supported, missing, or contradicted.
06
Use pass, review, and reject states
A score is easiest to operate when it supports three human-readable states. Pass means the public evidence supports the configured first-pass requirements. Review means a material field is ambiguous or missing. Reject means a hard campaign constraint is clearly not met.
Operators should record overrides. If a reviewer retains a low-score profile for a demonstrated skill or rejects a high-score profile for a campaign-specific risk, the note keeps the process explainable without pretending the rubric predicted the final choice.
07
What to verify after scoring
Open recent work, confirm product and customer fit, review brand-safety requirements, and verify that the public contact route is legitimate. Then discuss availability, rates, deliverables, rights, revisions, product access, and disclosure with the creator or representative.
Do not use the score as evidence of audience authenticity, expected response, creative quality, regulatory compliance, or paid-media performance. Those decisions need different evidence and, in some cases, specialist review.
A complete screening record should include the campaign brief, source profile, review date, reasons, exclusions, missing evidence, reviewer, and final status. Keeping those fields together makes the recommendation reproducible and lets the team improve requirements without claiming that one campaign decision applies to every future search.
Review the rubric whenever product behavior, provider fields, or campaign requirements change, and date material methodology updates.
08
Fit scoring is not audience-quality scoring
Two different things are called creator scoring, and buying the wrong one is a common mistake. Audience-quality scoring examines the follower base: fake-follower rates, audience age and country splits, engagement-rate benchmarks, and fraud signals. Tools such as HypeAuditor and Modash are built for that job, and a large influencer buy usually needs it.
Fit scoring examines the creator against one brief: does the public evidence show the right niche, country, language, content format, size band, and a usable contact route? UGC Agent does the second job and does not do the first. It publishes no audience-authenticity score, no fake-follower rate, and no engagement benchmark.
The distinction matters most for UGC. A paid UGC creator makes content the brand runs as an ad on the brand's own account, so the creator's own audience is often small and is not what the brand is buying. Audience metrics are close to irrelevant in that case, and evidence that the person can actually demonstrate a product on camera is what decides the shortlist.
Use audience-quality scoring when the creator posts to their own followers and the reach is the product. Use fit scoring when you are buying the content itself. Many teams need both, from two different tools, and should not expect one score to answer both questions.
Questions
What teams usually need to know.
What is the best way to score UGC creators?
Score them against the brief, not against their audience. For paid UGC the brand runs the content itself, so niche, country, content format, demonstrated product-demo ability, and a usable contact route decide the shortlist.
Does UGC Agent detect fake followers?
No. It publishes no audience-authenticity score, no fake-follower rate, and no engagement benchmark. Audience-fraud detection is a separate category of tool, such as HypeAuditor or Modash.
What is in the fit score?
A 15-point score built from public signals: UGC language in the handle, display name, or bio, public contact information, portfolio language and known portfolio hosts, and structured niche descriptions, checked against the campaign relevance requirements.
Can I see why a creator scored what it scored?
Yes. Every retained creator carries the reasons behind its number, and the reasons export with the row, so a reviewer can overrule the score and record why.
Does a high score mean the creator will accept the project?
No. The score is a first-pass screening aid on public evidence. It does not predict a reply, a rate, availability, creative quality, or campaign performance.
Can I change what the score rewards?
Yes. Saved preferences, one-run overrides, seed-creator analysis, niche requirements, and agency-email exclusions change which profiles survive screening.
Turn the brief into a creator list.
Try a restricted sourcing run free, then upgrade only when you need more visible or monthly creator leads.