In short
- Score every creator out of 100: audience authenticity 30, engagement quality 25, buyer fit 20, content craft 15, brand safety 10. Shortlist at 70 or more; below 60 is dropped.
- Work in three gates: a public screen worth 55 points, a creator-data check worth 38 and a conduct check worth 7. Stop when 70 is out of reach.
- Some findings end the process whatever the score: edited screenshots, a refusal to label, an undisclosed competitor deal in your window, or missing credentials for health and finance.
- No evidence scores zero. Date every screenshot, keep it with the sheet and re-vet before you rebook.
An influencer vetting checklist should end in a number. Score each creator out of 100 on five criteria: audience authenticity 30, engagement quality 25, buyer fit 20, content craft 15 and brand safety 10. Shortlist creators who reach 70, keep 60 to 69 on a reserve list and drop anyone below 60. This post is the sheet: 22 scored checks, three gates, instant-fail rules, a worked example and notes for different creator types. Copy the tables into a spreadsheet, or print this page and score by hand.
The 14-point checklist in our guide to how to find influencers in India is the quick version. This is the scored one, built on the Gully Score we apply to every creator before a brand sees a name. The sub-point split inside each criterion is our suggested split for a manual sheet; the five criterion totals and the 70 and 60 cut-offs are the Gully Score's own. For the channel as a whole, see the complete guide to influencer marketing in India.
How to award points
Every check has a maximum. Give full points for a clear pass, half for partial or uncertain, zero for a fail. No evidence scores zero. If a creator will not or cannot show the data, the benefit of the doubt does not come out of your budget. Write a one-line note beside every score so the next person knows why.
Three gates, and when to stop
The checks fall into three gates, in order of effort. Work through them in sequence and stop when the maths says the creator cannot reach 70.
| Gate | What it covers | Data needed | Max points |
|---|---|---|---|
| 1. Public screen | Content, comments, views, ads and followers you can see without asking | The public profile | 55 |
| 2. Creator data | Audience and reach, from the creator's own insights | Dated screenshots | 38 |
| 3. Conduct and paperwork | Delivery, conflicts and credentials | Messages, references and documents | 7 |
The stop rules come straight from the arithmetic. Gates 2 and 3 are worth 45 points together, so a creator with fewer than 25 points after gate 1 cannot reach 70. After gate 2, a creator needs gate 1 plus gate 2 of at least 63, because gate 3 adds at most 7. Stop there, and spare yourself the rest. Plan about 40 minutes per creator once the insights arrive: ten for gate 1, fifteen for gate 2 and fifteen for gate 3. That is a working estimate, not a benchmark.
The scorecard
Print this section. One column per creator works well in a spreadsheet.
A. Audience authenticity (30 points)
| # | Check | Gate | How to check | Max | Score |
|---|---|---|---|---|---|
| A1 | Audience cities and countries are plausible for the creator's language and content | 2 | Dated insights screenshot | 8 | |
| A2 | Follower growth over 90 days is gradual, or spikes are explained | 2 | Growth chart; ask about each jump | 6 | |
| A3 | Reach per post is a steady, plausible share of followers | 2 | Reach on the last 10 posts | 8 | |
| A4 | A sample of 30 followers looks like real people | 1 | Open 30 followers: photo, posts, names | 4 | |
| A5 | Screenshots are dated, consistent and match public numbers | 2 | Cross-check; ask for a live screen-share if doubtful | 4 |
B. Engagement quality (25 points)
| # | Check | Gate | How to check | Max | Score |
|---|---|---|---|---|---|
| B1 | Engagement rate is in line with the creator's tier | 1 | Engagement rate calculator on the last 12 posts against the tier ranges | 6 | |
| B2 | Median Reel views per follower is healthy; spikes are explained | 1 | Last 10 Reels, median not mean | 7 | |
| B3 | Comments are specific, on-topic and in the audience's language | 1 | Read 40 comments across 5 posts | 6 | |
| B4 | Saves and shares are in proportion to likes | 2 | Insights for the same Reels | 4 | |
| B5 | No pod signature: same commenters, early generic comments | 1 | First-hour comments on 5 posts; see engagement pods explained | 2 |
C. Buyer fit (20 points)
| # | Check | Gate | How to check | Max | Score |
|---|---|---|---|---|---|
| C1 | Audience age, gender and city overlap with your buyer | 2 | Insights against your buyer brief | 8 | |
| C2 | Language matches the way your buyer speaks | 1 | Captions and Reels | 4 | |
| C3 | Organic content already touches your category | 1 | Last 15 posts | 4 | |
| C4 | Comments show purchase intent | 1 | Questions about price, size, where to buy | 4 |
D. Content craft (15 points)
| # | Check | Gate | How to check | Max | Score |
|---|---|---|---|---|---|
| D1 | Hooks hold attention in the first three seconds | 1 | Watch 10 Reels | 4 | |
| D2 | Clear audio, light and framing | 1 | Same Reels | 3 | |
| D3 | Sponsored posts perform close to organic ones | 1 | Last five sponsored against last five organic | 5 | |
| D4 | Delivers on time and takes feedback; brands would work with them again | 3 | Reply speed; a reference or two | 3 |
E. Brand safety (10 points)
| # | Check | Gate | How to check | Max | Score |
|---|---|---|---|---|---|
| E1 | Past paid posts carry a clear, upfront label | 1 | Open five past sponsored posts | 3 | |
| E2 | No competitor deal inside your exclusivity window | 3 | Search recent sponsored posts; ask | 2 | |
| E3 | Comment section tone and past controversies are acceptable | 1 | Scan comments; search the creator's name and handle | 3 | |
| E4 | Credentials and claims discipline for health and finance | 3 | Ask for proof; read how claims are made | 2 |
The gate totals are 55, 38 and 7, and the criterion totals are 30, 25, 20, 15 and 10, which is 100.
Instant-fail rules
Some findings end the process whatever the score.
| Finding | Why it ends the process |
|---|---|
| Screenshots edited, and the creator refuses a live screen-share | Dishonest data makes every other number unusable |
| Reel likes higher than plays | This should not happen on real viewing, so likes are likely bought |
| The creator will not use a disclosure label | The brand shares responsibility for the post under ASCI's guidelines |
| An undisclosed competitor deal inside your exclusivity window | A conflict you cannot cure |
| Health or finance advice without the required credentials | ASCI's Addendum II requires qualifications stated upfront, and SEBI registration for stocks and investments |
| Two requests for insights unanswered | Professional accounts have the data; refusal is itself a signal |
Brands that SEBI regulates should also note SEBI's October 2024 circular, which bars regulated entities from associating with people who give securities advice without registration. This is general information, not legal or tax advice.
Verdict bands
| Score | Verdict | Next step |
|---|---|---|
| 80 to 100 | Shortlist | Send to whoever approves budget with the sheet |
| 70 to 79 | Shortlist, with a note | Write down why it is not higher, and make any open question a condition of booking |
| 60 to 69 | Reserve | Book only after a small paid trial, or if a shortlisted creator drops out |
| Below 60 | Drop | Keep the sheet; do not re-audit the same account next quarter |
A worked example
Adjust the sheet for the creator
| Creator type | What changes |
|---|---|
| Nano | Thinner history, so medians swing. Keep A1, B3 and a small trial; accept fewer data points |
| Micro and mid-tier | Use the full sheet. Add the full 12-post engagement calculation |
| Health and finance | E4 becomes an instant fail if credentials are missing, per ASCI's Addendum II |
| UGC-only | Skip A, B and C; use the portfolio scorecard in how to find UGC creators in India |
| Regional language | Have someone fluent in the language read the comments for B3 and C2 |
| Local business | Weight A1 heavily; city share matters more than total reach, as in how to find local influencers in your city |
Keep the evidence
Save each creator's screenshots with the date visible, in a folder named by creator and date, and keep the sheet beside them. Ask creators to warrant in the contract that the data is accurate and unedited; our influencer contracts in India guide has the clause. Re-vet before every rebooking, after any sudden follower jump and at least every 90 days on a long deal.
Request the same data from every creator, in the same words. The messages are in influencer outreach message templates, what to do with the screenshots is in how to read influencer audience insights, and how to read what a creator sends first is in influencer media kit: what to look for. For small accounts, our guide to finding micro-influencers shows where to look.
ASCI's numbers explain why E1 exists: in FY2025-26 the regulator processed 1,609 influencer ads and 97.3% needed modification.
Mistakes with checklists
- Scoring without evidence. A score with no note cannot be audited.
- Changing the weights per creator. Fix the weights first, then compare.
- Letting a high buyer-fit score overrule authenticity. Great fit with a weak audience is the most common way budgets are wasted.
- Skipping the stop rules. They exist to save time.
- Auditing once. Audiences change.
We run this process for every creator we propose, across our Instagram influencer marketing work and for brands that want it done for them. Send us a brief and the plan, scored and with names, returns within 48 hours of the brief call.
Sources
- Indian consumers and influencers: coverage of a 2025 Kantar-backed study of consumers and brands, Campaign India (12 Jun 2025)
- ASCI annual complaints report 2025-26, ASCI (May 2026)
- ASCI half-yearly complaints report 2025-26, ASCI (Nov 2025)
- ASCI Guidelines for influencer advertising in digital media (including Addendum II on health and financial influencers), ASCI (accessed 31 Oct 2026)
- SEBI circular on association of regulated persons with persons giving unregistered securities advice, SEBI (22 Oct 2024)
Frequently asked questions
What should an influencer vetting checklist include?
Audience location, age and gender, reach per post, engagement rate, Reel views against followers, comment quality, buyer fit, content quality, disclosure habit, competitor conflicts, comment-section tone and credentials where the category needs them. A good checklist ends in a score and a cut-off, so every creator is judged the same way.
How long does it take to vet one influencer?
Plan on about 40 minutes per creator once their insights arrive: ten for the public screen, fifteen for the data check and fifteen for content, conduct and paperwork. That is our working estimate, not a published benchmark. A quick public screen first saves time on creators who fail it.
What score should an influencer need to be shortlisted?
On the Gully Score, creators need 70 or more out of 100 to be shortlisted, and anyone below 60 is dropped. Between 60 and 69 they can sit on a reserve list. Use the same cut-offs in your own sheet so that scores from different niches stay comparable.
Do I need to vet nano influencers as carefully?
Yes, with lighter data. Nano accounts have thinner history, so medians swing more. Keep the audience location check and the comment review, and run a small paid or barter trial before a larger commitment. The fee is smaller, but wrong-audience posts waste product and time just the same.
When should I re-vet a creator?
Before every rebooking, after any sudden follower jump, and at least every 90 days for an ongoing deal. Audiences change, and a creator who scored 82 last quarter may not now. Keep the old screenshots so you can compare. Our fake followers guide explains what to look for.
Can I use this checklist for UGC creators?
Only partly. If the video runs from your account, audience criteria do not apply and craft is the whole job; use the portfolio scorecard in how to find UGC creators in India. If the creator posts it or you run it from their handle, use this full sheet.