Every UGC creator looks good in a portfolio. That is what portfolios are for. The trouble shows up after the invoice clears: the video lands four days late, the hook is the one they used last month for a protein powder, the product sits on a shelf behind them instead of being used, and the raw files you assumed were included cost extra. You have paid for something you can't run.
It happens more than brands like to admit. In the Foxwell Founders and Motion "State of UGC 2026" report, as reported by Net Influencer, 57% of the marketers surveyed said they have no clear read on whether their UGC creator spend pays back. Some of that is measurement. A lot of it starts earlier, when someone picked a creator off a grid of thumbnails and hoped.
Learning how to vet UGC creators is the cheapest fix in the whole process, because it happens before any money moves. Below is the filter we'd give any operator holding a shortlist. Seven checks go into a doc. A ledger explains what each red flag usually means. Then a small paid trial settles the one question no checklist can answer. The criteria come from the creator section of Spark's own Marketer's Blueprint (authenticity, engagement, a performance track record, delivery, responsiveness), tightened with what has changed by 2026.
The 7-check UGC creator vetting scorecard (copy it)
Run it once per creator. Every check is pass or fail, and each comes with a test you can do from your desk or inside one message. Three checks are gates: fail any of them and the creator is out, however good the reel looks. If I could keep only one, it would be check 5. A creator who asks nothing about your customer before quoting will make the same video for you that they made for the last brand.
UGC CREATOR VETTING SCORECARD
Creator: ____________ Found on: ____________ Date: ________
Mark each check PASS or FAIL. A FAIL on any [GATE] ends the review.
1. SOUNDS LIKE A REAL PERSON
Test: watch 3 full videos with the sound on. Not the thumbnails.
Pass: talks like they're telling a friend; the product is picked up and used.
Fail: reads a script at you, or every video is the same glossy set.
2. RANGE BEYOND ONE TEMPLATE
Test: ask for 3 pieces in different formats (talking head, demo,
voiceover over b-roll).
Pass: different openings, settings and pacing.
Fail: all from one platform, same hook shape, same corner of the room.
3. A PAID TRACK RECORD
Test: ask "which of these ran as a paid ad, and what did you learn
from how it did?"
Pass: names at least one ad that ran with spend, or has handed over
a Spark Ads code or partnership ad access before.
Fail: a vague answer, or results they can't explain.
New creator with no ads yet: mark N/A and let the trial decide.
4. ENGAGEMENT THAT LOOKS HUMAN (if they post their own content)
Test: open their last 10 posts and read the comments.
Pass: comments react to what was actually said; posting is recent.
Fail: rows of short identical praise, a sudden follower jump,
or a feed left alone for months.
5. QUESTIONS BEFORE QUOTES [GATE]
Test: send a one-paragraph brief and wait.
Pass: they ask about your customer, the main benefit, the objections
or where the ad will run.
Fail: a rate card, "when do you need it?", nothing else.
6. FOLLOWS A REAL INSTRUCTION, ON TIME [GATE]
Test: plant one specific instruction in the brief
(e.g. "open with the product already in your hand").
Pass: replies inside your working window and acknowledges it.
Fail: slow now (it won't speed up after payment) or ignores it.
7. TERMS IN WRITING [GATE]
Test: ask for revisions, raw files, usage term, ad permissions and
payment terms in one written reply.
Pass: clear answers, a deposit or milestone split, happy to grant
partnership ad or Spark Ads access.
Fail: 100% upfront with nothing written down, "use it however you
like" on rights, or a no on ad permissions.
RESULT
Invite to the paid trial when all three [GATE] checks pass and no more
than one of checks 1 to 4 fails. N/A on check 3 is not a fail.
How do you run the scorecard on a shortlist?
Order matters, because half the checks need nothing from the creator and the other half can be bundled into a single message.
- Checks 1, 2 and 4 only need the portfolio and the public profile. Do them before you contact anyone, and a long list gets short quickly.
- Next, send the short brief with your planted instruction, plus the terms questions from check 7, in the same note. Their reply tests checks 5, 6 and 7 at once. If you haven't written a brief yet, start from the UGC brief template.
- Ask about paid history (check 3) last. It lands better once they've seen the brief, because the answer tells you whether they think about ads or just about posting.
- Log every result: one row per creator, the date, the fails, a note. A creator who passed in March and missed a deadline in May should carry that note into the next batch.
One thing the scorecard deliberately leaves out is follower count. In CreatorIQ's 2026 State of Creators research, reported by Marketing Dive, brands ranked suitability as their top consideration when choosing creators and put follower count last of eight. For UGC ads that ranking makes even more sense. Your ad shows to your ad account's audience, not to the creator's followers, so their audience size barely matters except as a check that the account is real.
The same research found 62% of creators say content creation isn't their main source of income. That is why check 6 is a gate. Plenty of good creators film around a day job, and that is fine, as long as they tell you their turnaround honestly and then hit it.
UGC creator red flags: the ledger
The scorecard catches most problems. These are the specific patterns that tend to sit behind a fail, with what each one usually means and what to do about it.
| Red flag | What you'll see | What it usually means | Your move |
|---|---|---|---|
| Portfolio all from one platform | Every piece is the same native talking head, same setup, same opening line shape | Template-bound: one format, repeated for every brand | Ask for one piece outside that format in the trial. If they won't, pass |
| No questions about your audience or benefits | A rate card by return and zero curiosity about the product | A volume churner filming the same script shape for everyone | Gate fail on check 5. Move on |
| 100% upfront, nothing in writing | "Full payment before I start" with no agreement attached | You carry all of the delivery risk with someone you've never worked with | Offer a deposit with the balance on delivery, or pay through a platform that holds funds until the content arrives. Walk if refused |
| Repeated short, bot-style comments | Dozens of "so good!!" from accounts with no posts of their own | Bought engagement. The FTC's 2024 rule bans buying or selling fake indicators of social influence, such as bot-generated followers or views | You aren't buying their audience for UGC ads, so it isn't fatal on its own. It is an honesty signal: fail check 4 |
| Their own socials are abandoned | Last post months ago, dead link in the bio | A side gig that lapsed, or a portfolio that is older than it looks | Ask when the newest portfolio piece was filmed |
| Dodges ad permissions | Vague when you mention partnership ads or a Spark Ads code | You can't run the video from their handle, which is often where native-looking ads do best | Settle it before you pay. On TikTok the creator generates a post code from the video's Ad settings |
| Vague usage rights | "You can use it however you want" | Nothing you could point to if a dispute came up at renewal | Get the term, channels and paid use in writing |
| Unsure about disclosure | Doesn't know when their own posts need an ad label | A risk if they post organically for you | Brief the disclosure in writing and check the first post |
| Rate far below market, unlimited revisions | A quote that looks too good next to everyone else's | Something gives later, usually quality or the deadline | Benchmark it against realistic creator rates before you book |
Two of those rows have a legal side worth knowing. The FTC's final rule on fake reviews and testimonials covers fake social media indicators, which is why bought followers are more than a vanity problem. And the FTC's Disclosures 101 for social media influencers is clear that a free or discounted product counts as a material connection that needs disclosing. Pure UGC that only runs from your brand's page sidesteps most of this, but the moment a creator posts for you on their own account, it applies.
For the rights row, the full list of what to pin down is in the UGC usage rights checklist. For the price row, compare against the real all-in creator rates for 2026. For ad permissions, Meta's version is covered in the partnership ads and whitelisting guide, and TikTok's own help centre shows creators generating a Spark Ads post code from the three-dot menu on the video.
The paid trial: the vetting step that uses real spend
The scorecard tells you who is safe to pay. It can't tell you who is good at ads, because nobody knows that until a video has run against a real audience. That is what the trial is for. The Blueprint's advice is to test at least three creators, and we'd hold to it. With one, you have nothing to compare against. With three on equal budgets, you can tell whether a result came from the creator or from the concept.
- Three finalists, one brief. Same product, same angle, same planted instruction. Only the creator changes, so the difference you see belongs to the creator.
- A small, paid deliverable. One short video plus two alternative hooks filmed in the same session. Pay their normal rate. Asking for free "test" videos mostly attracts the creators with nothing else booked.
- Grade the raw files before they touch the account. Was it on time? Was the planted instruction followed? Is the product actually used on camera? Is the audio clean? Are the hooks genuinely different, or the same line reworded twice?
- Run a fair test. Same campaign, same budget per ad, same window. Read hook rate and hold rate first, because they settle sooner than CPA. If those terms are fuzzy, the creative metrics glossary defines them.
- Rebook the winner, keep the runner-up warm. Add notes on everyone to your log. The creator who came second on this product may be the right face for the next one.
Then write the brief for the winner properly. A good creator with a thin brief still produces a thin ad, so the next stop is how to brief a UGC creator.
What vetting really costs, batch after batch
Add it up. Desk checks per creator, a message thread per creator, trial fees for three, ad spend on the test, grading time, the log. Then do it again next batch, because winning ads fatigue and you need fresh faces and fresh angles to replace them. For a brand buying a handful of videos a quarter, that is manageable, and you should run this filter yourself. For a brand that needs new creative every couple of weeks, the vetting quietly becomes a part-time job, which is the hidden line in DIY UGC versus done for you.
That workload is the part Spark takes on. We cast creators from our vetted network to your brand and audience, brief them, and run the shoots for performance in-feed, drawing on more than 5,000 ads made. If you would rather keep casting in-house, the scorecard above is yours to keep. If you would rather stop re-vetting every batch, have a look at how our plans work or browse more guides in Resources.
Frequently asked questions
How do you vet a UGC creator before hiring them?
Check the portfolio for natural, phone-native delivery and range across formats, ask which videos ran as paid ads, and read the comments on their own posts for bot-style engagement. Then send a short brief: strong creators ask about your audience and benefits, reply promptly, and put revisions, raw files, usage rights and payment terms in writing. Finish with a small paid trial across at least three creators.
What are the biggest red flags when hiring a UGC creator?
No questions about your customer or product benefits, a portfolio that repeats one template from one platform, a demand for 100% payment upfront with no written terms, rows of short generic comments that look automated, abandoned social accounts, and reluctance to grant partnership ad or Spark Ads permissions.
Should you pay a UGC creator upfront?
A deposit is normal and fair, because creators carry risk too. Full payment before any work, with no written agreement, to someone you have never worked with, puts all of the delivery risk on you. Split it into a deposit and a balance on delivery, or pay through a platform that releases funds when the content arrives.
How many UGC creators should you test at once?
At least three on the same brief. One creator gives you nothing to compare against, while three running on equal budgets show you whether a result came from the creator or from the concept.
Start with your shortlist
Paste the scorecard into a doc, run the desk checks on everyone you've saved, and send the gate message to whoever survives. By Friday you should know which three are worth a paid trial. When you want the casting handled for you, start a project with Spark.