Influencer Vetting Checklist
A 15-point influencer vetting checklist for crypto, iGaming and fintech teams: audience quality, engagement fraud, brand safety and past promos.
Head of Marketing at Oznaka
Influencer vetting is the due diligence you run before paying a creator. You check that the audience is real, the engagement isn't farmed, past sponsored posts actually performed, and the creator is safe for your brand and your regulator. In crypto, iGaming, and fintech, add two more checks: disclosure history and any record of promoting projects that later failed.
Most vetting guides are written for skincare and fashion brands on Instagram. They ask whether the creator's look matches yours and whether the audience is the right age. When we vet creators for crypto campaigns, the first questions are different. Is the influencer audience really interested in products like yours? How did their past content on the same topic perform? What did their audience say about that? Do they have organic reach to the right people?
In our practice, we had a case with a prediction markets project where one influencer who wasn’t vetted initially performed better than the one they came with and asked to include. What was the reason? Unique audience. The one they wanted to include wrote about prediction markets. The one we suggested was from the linked niche – betting – and had less overlap with other authors. We did this test and found out this played well – we talked to the new audience with the same motivation and needs.
Loading embed…
The 15-point influencer vetting checklist
Mark each check as pass, flag or fail. A simple rule: drop a creator after any fail on checks 1, 4, 12 or 14, or after three flags in total. The other fails are worth a conversation before you decide.
Audience (checks 1–5)
| # | Check | How to do it | Red flag |
|---|---|---|---|
| 1 | The audience is real | Scroll through 30 random followers and run a fake-follower check | Most followers have no photo, a handle full of digits, no posts, and joined in the same month |
| 2 | Growth has a reason | Open the follower history and match each spike to a viral post, a repost or a press mention | Jumps of thousands in a day with nothing behind them, then a slow decline |
| 3 | The audience is in your markets | Check reply languages, the "About this account" country of commenters, and the creator's own analytics | Most of the audience sits in countries where your product isn't available or licensed |
| 4 | The audience is old enough | For iGaming and trading products, ask for the age split from native analytics | More than 25% under 18 (the UK gambling threshold), or the creator won't share the data |
| 5 | The audience cares about your niche | See whether followers also follow projects in your category | A "crypto" creator whose followers mostly follow giveaway and airdrop-farming accounts |
Checks 1 and 2 are covered in depth in our guide on how to spot fake influencers.
Engagement (checks 6–9)
| # | Check | How to do it | Red flag |
|---|---|---|---|
| 6 | Engagement fits the tier | Calculate the median engagement rate over the last 10–20 posts and compare it with creators of the same size and niche | Far below peers, or far above them with no obvious reason |
| 7 | Replies come from real people | Read the replies on 3–5 posts and open a few of the reply profiles | Generic replies, and the same 30–40 accounts under every post |
| 8 | Reach is organic | Compare views, likes and replies on the same posts | 300K views with 40 likes, or big numbers only on sponsored posts |
| 9 | Performance is stable | Judge the median post, not the best one | One viral post carries the whole average |
On X, replies and reach are where most fraud hides. See engagement farming on X for how raid groups work, and our Twitter (X) audit for reading views against engagement.
Content and brand safety (checks 10–12)
| # | Check | How to do it | Red flag |
|---|---|---|---|
| 10 | Past content is safe | Scan 12 months of posts, replies and reposts. Search the creator's name with words like "scam" and "rug" | Harassment, public fights with projects, or promoting schemes that were later exposed |
| 11 | They know the topic | Read their own threads and replies. Look for original takes rather than reposted alpha | They post about whatever token is trending that day, with no consistent focus |
| 12 | They disclose paid posts | Find past sponsored posts and check for a clear ad label | Paid campaigns you can identify that carry no label at all |
Commercial history (checks 13–15)
| # | Check | How to do it | Red flag |
|---|---|---|---|
| 13 | Past sponsored posts performed | Ask for a screen recording of native analytics for their last 3 sponsored posts, and compare with organic posts | They refuse, or sponsored posts get a fraction of normal reach |
| 14 | Promoted projects held up | List the last 10 projects they promoted and check where each one is now | A pattern of tokens that collapsed soon after the promo |
| 15 | No conflicts | Ask about current competitor deals, token allocations, affiliate links and exclusivity | They promoted your direct competitor this month, or hold an allocation they can sell at launch |
Check 14 matters more than it looks. A Harvard Business School study of 36,000 tweets from 180 crypto influencers found that promoted tokens were down 6.53% on average 30 days later and 18.9% after 90. If a creator's past calls cost their followers money, those followers are less likely to act on the one you're paying for.
The influencer vetting process, step by step
Run the checks in this order, so the slow ones only happen for creators who survive the fast ones.
- Start from the brief. Write down the markets, the audience, the budget per creator and your red lines (for example, no audience under 18) before you look at a single profile.
- Run a two-minute filter. Compare follower count with the quoted price, glance at the last 20 posts, and check the free X Score on X. Cut the obvious misfits here.
- Run the 15 checks. Audience and engagement (checks 1–9) are where a tool saves the most time. Content and commercial history (10–15) stay mostly manual.
- Ask the creator for data. Request a screen recording of native analytics: audience age and countries, plus the last three sponsored posts. Ask which deals they're running now.
- Score and decide. Fill in the scorecard and write one line on why each dropped creator was dropped, so nobody re-checks them next quarter.
- Put it in the contract. Disclosure on every post, your right to request analytics, no deleting the post for an agreed period, and part of the fee tied to delivery.
- Re-check before you renew. A creator who passed six months ago may have bought followers or picked up a competitor deal since.
Manual vetting vs a scan: how long it takes
We timed both on 30 creators from a real shortlist.
| Part of the checklist | By hand | With an Oznaka scan |
|---|---|---|
| Audience and engagement (checks 1–9) | 2-3 minutes per profile if you’re doing this often enough and know where to watch | Less than 2 minutes from your brief to results agent collects for you |
| Content and brand safety (checks 10–12) | Not always visible at all | Agent can run this check and find signals |
| Commercial history (checks 13–15) | Quick manual check plus the creator's reply | Same, needs the creator's data |
| Total per creator | Optimistic scenario: 10 minutes per one profile | 3 minutes to collect the 200 best-fit creators |
| A shortlist of 30 creators | 6 hours | Depends on your intention, but typically we hear a feedback that actionable list can be built in 1-2 hours |
The gap is widest on checks 1–9, because a profile can't show you how much of the reach was boosted or who the people replying really are. That's the part Oznaka (oznaka.ai) scans on X: organic reach, organic replies, niche match, a price benchmark and an overall Fit Score.
Crypto-specific red flags
Crypto creators, or KOLs as the industry calls them, get paid in ways that generic vetting tools don't track. Three patterns deserve their own check.
Paid and unpaid shilling
Some crypto creators shill projects in almost every post. It doesn’t necessarily mean they don’t have a real audience, but the ability to drive actions with such posting is questionable. If you see a KOL whose profile looks like a billboard, that’s a sign to stop and check it more precisely.
Giveaway KOLs
3 SOL to 3 winners in 24 hours. People participating in such events don’t need your product, so you most likely don’t need to pay those who host them.
Fake screenshots
Some creators writing about crypto trading can fake their stats and prove this with fake screenshots. Run checks, do due diligence, stay curious. Reputational risks from working with such authors outperform potential benefits.
Deleted posts
Promoting token launches or anything token-related, you can notice that some creators remove promo posts once a token drops or promo period ends. To spot that, look for quote posts of theirs that now show "This post is unavailable".
Influencer vetting tools compared
No single influencer vetting tool covers all 15 checks. Most cover audience and engagement. The gaps are brand safety and, almost everywhere, a creator's commercial history. Pick by the platform you buy on and the checks you can't do by hand.
| Tool | Platforms | Checks it covers | Best for | Limits |
|---|---|---|---|---|
| Oznaka | X | 1, 3, 5–9 and a price benchmark | Crypto and fintech teams buying on X | X only for now; brand safety and commercial history stay manual |
| X Score | X (Chrome extension) | A quick read on 1, 7 and 8 as one 0–1000 score | The two-minute filter | One number, no breakdown |
| HypeAuditor | Instagram, YouTube, TikTok, Twitch, X, Snapchat | 1–3, 6–8, demographics | Consumer brands on Instagram and YouTube | Paid; only accounts over 1,000 followers |
| Modash | Instagram, TikTok, YouTube | 1–3, 6, 13 (paid vs organic posts, brand-collab history) | Instagram and TikTok programs | No X coverage |
| Kroll | All major platforms | 10 and ongoing monitoring, reviewed by investigators | High-profile talent and large contracts | A service, not a self-serve tool; no audience metrics |
| Manual scorecard | All | All 15 | Fewer than 10 creators | Slow; can't see boosted reach or who repliers are |
Brand safety deserves more attention than most teams give it. Kroll says about one in four influencers it has vetted carried significant risk. And if all you need is a quick answer on a single X account, start with our checklist for spotting bot accounts on X.
Free influencer vetting scorecard
You don’t need a complex tool for that. We put the whole checklist below, so you can copy and reuse it. One row per creator, one column per check, and the decision fills in on its own. No matter if you run this in a spreadsheet or in your AI agent’s setup.
| Column | What goes in it |
|---|---|
| Creator, platform, quoted price | The basics from your shortlist |
| Checks 1–15 | Pass, flag or fail from a dropdown |
| Critical fails | Counted for checks 1, 4, 12 and 14 |
| Total flags | Total flags you found |
| Decision | Go, talk first, or drop, based on the rule above |
| Why dropped | One line, so nobody re-checks this creator next quarter |
| Date checked | So you know when to re-check before a renewal |
FAQ
Before you pay
Vetting influencers properly takes time, so spend it where the money goes. Run the two-minute filter on the full list and save the 15 checks for whoever passes. Keep every result in the scorecard, because next quarter's shortlist starts there.
On X, Oznaka covers the audience and engagement checks in one scan.
Find creators who fit your next campaign.
Start with your audience, your niche, and the outcome you want to achieve.
Find creators