In short
- Run five gates first (label, claims, product facts, rights, conflicts). A gate failure is fixed before any score matters.
- Then score ten criteria from 0 to 3, for 30 points. Hook, message, product in use, voice, proof, craft, language, call to action, pacing and reuse.
- Score the live post out of 15 against the creator's own organic Reels, not a brand-wide average. Compare draft and live scores to learn what failed.
- Give feedback as one ranked list with timestamps and reasons, and keep to the revision rounds in the contract.
To evaluate influencer content quality, check five gates and then score ten criteria. The gates are pass or fail: the disclosure label, the claims, the product facts, the rights and any conflicts. If any gate fails, the draft goes back before anything else is discussed. The ten criteria are scored from 0 to 3 for a total of 30 points: hook, message, product in use, creator voice, proof, craft, language, call to action, pacing and reuse. After the post goes live, add a 15-point block that compares its results with the creator's own organic content.
Our influencer vetting checklist scores creators before you book them, and the Gully Score gives content craft 15 of its 100 points, judged from past work. This rubric judges the work a creator makes for you, which is better evidence for the next booking. In a 2025 Kantar-backed study covered by MediaNews4u, marketers rated content quality (36%) and engagement rate (39%) above reach as indicators of performance, so a shared way to score quality pays off. For the wider channel, see the complete guide to influencer marketing in India.
Gates first: five checks that are pass or fail
Gates come before scores because a charming draft with a missing label still needs fixing. ASCI processed 1,609 influencer ads in FY2025-26 and 97.3% needed modification, so assume your draft has a gate problem until you have checked.
| Gate | What to check | A pass looks like |
|---|---|---|
| 1. Disclosure | A label in the right place and for long enough | See the timing table below |
| 2. Claims | Every claim is in the brief and one the brand can substantiate | Nothing added on camera; no health or finance advice without the credentials ASCI requires |
| 3. Product facts | Product name, price, offer, availability, how it is used | Matches the brief, including the code and link |
| 4. Rights | Music, footage and any third-party brand are cleared for a branded post | The creator confirms the audio carries the rights the contract needs |
| 5. Conflicts | No rival brand visible, no unsafe use of the product | A frame-by-frame scan finds none |
ASCI's influencer guidelines ask for the label to be upfront and prominent, not only in a bio, not at the end of a post, not behind a "more" click and not buried in hashtags. The timing depends on format:
| Format | Where ASCI says the label goes |
|---|---|
| Video of 15 seconds or less | On screen for at least 3 seconds |
| Video over 15 seconds and under 2 minutes | On screen for one third of the video |
| Video of 2 minutes or more | For the whole section that mentions the brand |
| Image or video without a caption, such as Stories | Superimposed on the visual |
| Live stream | Announced at the start and end |
| Audio | Announced at the start, the end and around breaks |
Permitted labels include Advertisement, Ad, Sponsored, Collaboration, Partnership and a platform's own disclosure tool. This is general information, not legal or tax advice. The full rules are in our ASCI influencer guidelines explainer, and wording by platform is in influencer disclosure hashtags in India.
The ten criteria
Score each from 0 to 3. A 3 is clearly good, 2 is acceptable, 1 is weak and 0 is absent or harmful. Write a one-line note beside every score so that the next reviewer can see why.
| # | Criterion | A 3 looks like | A 0 or 1 looks like |
|---|---|---|---|
| 1 | Hook | The first 3 seconds give a reason to stay: a problem in the viewer's words, a surprise, a question | A logo, a greeting or "so today I am..." |
| 2 | Message | One main message, said once, in plain words | Three messages, or none you can repeat |
| 3 | Product in use | The product is used on screen: texture, size, result, a real moment | Held up to the camera and described |
| 4 | Creator voice | Sounds like the creator's own posts | Reads like a brand script |
| 5 | Proof | A concrete detail the brand can back: a price, a wear time, a comparison | Adjectives only |
| 6 | Craft | Clear sound, light, framing and subtitles; sensible pacing | Hard to hear, text hidden behind app buttons |
| 7 | Language | Natural in the audience's language or mix of languages | A word-for-word translation of the brief |
| 8 | Call to action | One clear next step, with the right code, link or tag visible | No action, or the wrong code |
| 9 | Pacing | The product appears before viewers typically drop off | The product appears in the last seconds |
| 10 | Reuse | Clean first frame, vertical, text inside safe areas; rights agreed for ads | Branded overlay that blocks reuse, or no rights |
The pacing check connects to watch time. If average watch time on the creator's recent Reels is 9 seconds and your product appears at second 14, most viewers will leave before they see it. The method is in how to measure Instagram Reel performance.
Reading the total
| Score (of 30) | Verdict | Next step |
|---|---|---|
| 24 to 30 | Approve | Pass to the brand's approver with the notes |
| 18 to 23 | Approve with edits | Send a ranked list of changes; check the edited cut |
| 12 to 17 | Rework | Use a revision round, and ask whether the brief was clear |
| Below 12 | Reshoot or release | Check the contract for remedies |
These bands are our suggestion. Two reviewers should score 5 drafts independently before you rely on them, and compare. If they differ by more than a few points on average, the anchors need rewriting.
Giving feedback on a draft
Send one consolidated list per revision round. Rank it: must fix (gates and claims), should fix (scores of 0 or 1) and could fix (taste). Quote a timestamp, name the problem and give the reason. "At 0:19 the product first appears; viewers usually leave by then" gives the creator a problem to solve. "Make it more premium" does not.
Two rounds of revisions are standard in our contracts, and our influencer campaign brief template shows how to set expectations for them. Agree a review turnaround in the brief too, because a draft that waits three days costs a posting slot. Approve the concept or script before the shoot, so that the rough cut only needs craft notes.
Scoring a live post at day 7
A live post has two jobs: it should be good, and it should perform against what the creator normally gets. Add a 15-point block, scored 0 to 3 on each of five items, using the creator's organic posts as the baseline.
| Item | 3 | 2 | 1 | 0 |
|---|---|---|---|---|
| Delivery: views against the creator's median | At or above | Up to 20% below | 20% to 50% below | More than 50% below |
| Hold: completion ratio against their organic posts | At or above | Up to 20% below | 20% to 50% below | More than 50% below |
| Response: saves and shares per 1,000 reached | At or above | Up to 20% below | 20% to 50% below | More than 50% below |
| Comments: questions and purchase intent | More than usual | About usual | Fewer | Mostly spam, emoji or complaints |
| Compliance: label, link and code in place | All correct | A minor fix | Label late | Label missing, or edited after approval |
The first three items use the ratios in our Reel performance guide. The comment item needs a hand-coded sample, as explained in social listening for influencer campaigns. The bands are our suggestion, not published thresholds, and we found no India benchmark for any of them.
What the two scores tell you together
| Draft score | Live score | Likely reading |
|---|---|---|
| High | High | Rebook; the creator and the brief both worked |
| High | Low | Check buyer fit and audience, not the content; see how to read influencer audience insights |
| Low | High | Do not assume it repeats; the creator's audience may have carried a weak draft |
| Low | Low | Drop, or rework the brief before rebooking |
A worked example (made-up numbers)
Adjusting the rubric by format
| Format | What changes |
|---|---|
| Instagram Story set | Score each frame's hook; add exits and forward taps from the creator's insights; label on the visual |
| YouTube integration | Pacing becomes where the integration sits in the video; check the retention graph before and after it |
| YouTube Short | Hook and the first 3 seconds carry more weight; check audio rights for any music |
| LinkedIn post | The hook is the first lines before "see more", and the label belongs there too |
| Photo or carousel | Replace pacing with slide order; the message must land by slide two |
| UGC for your own ads | Skip voice and audience items; add first-frame strength and the ability to cut into variants |
Mistakes in evaluating content
- Scoring after you see the numbers. Strong results make a weak draft look better. Score drafts before posting.
- Using one scorer. Two scores, compared, reveal unclear anchors.
- Treating taste as a gate. Gates are label, claims, facts, rights and conflicts. Everything else is a score.
- Overriding the creator's voice. Fix facts strictly and leave the idiom.
- Comparing with a brand-wide average. Compare with the creator's own organic posts.
- Not keeping the scores. Put them in the sheet, and in the Posts tab of an influencer marketing dashboard, so the next shortlist can use them.
If you want drafts checked for message, claims and disclosure before they reach you, that is a standard step in how we run campaigns; see how we work or send us a brief. Our Instagram influencer marketing service shows where this fits in a campaign.
Sources
- ASCI annual complaints report 2025-26, ASCI (May 2026)
- ASCI Code and guidelines, including the influencer advertising guidelines, Advertising Standards Council of India (undated, accessed 31 Oct 2026)
- India's influencer marketing industry set to grow 25% in 2025, finds WPP and Kantar report (engagement rate and content quality as performance indicators), MediaNews4u (11 Jun 2025)
- Indian consumers and influencers: coverage of a 2025 Kantar-backed study of consumers and brands, Campaign India (12 Jun 2025)
Frequently asked questions
How do you evaluate the quality of influencer content?
Check five gates first: disclosure label, claims, product facts, rights and conflicts. Then score ten criteria from 0 to 3: hook, message, product in use, creator voice, proof, craft, language, call to action, pacing and reuse. Compare the score with how the post performed against the creator's own organic content. Our influencer vetting checklist covers the earlier step of choosing creators.
What should I check first in an influencer's draft?
The gates. Is the disclosure label present, upfront and long enough on screen? Is every claim in the brief and one your brand can substantiate? Is the product, price and offer correct? Are music and footage cleared? Is any rival brand visible? Fix gate failures first, because they carry legal and brand risk that a good hook cannot offset.
How many revision rounds should I allow?
Two rounds is standard in our contracts, and it should be written into yours. Send one consolidated list per round, with must-fix items first. A third round usually means the brief was unclear, not that the creator is difficult. Our guide to influencer contracts in India covers revision clauses.
How do I give feedback without killing the creator's voice?
Fix facts and claims strictly, and give notes on style as problems to solve, not instructions. Say 'viewers will leave before the product appears at 0:19' rather than 'make it more energetic'. Quote timestamps, give the reason, and leave the idiom to the creator.
Does high engagement mean the content is good?
No. Engagement can come from the creator's audience loyalty, a giveaway or a pod, not from the quality of the branded post. Judge the branded post against the creator's own organic posts, and use the engagement rate calculator for the screening step only.
How do I score content in a language I cannot read?
Use a native reviewer for language, tone and claims, and score the rest from the visuals and a summary. Ask for an English summary with each claim flagged, as in how to find regional language creators. You approve the claims, and the reviewer approves the idiom.