How to Measure Influencer Campaign ROI
Reach is the easiest number to collect and the least useful. Here is what to track instead, and how to attribute sales without lying to yourself.
Israni Talent Hub Editorial26 Oct 20263 min read
Most influencer campaigns in India are reported on reach, because reach is the number the platform hands you for free. It is also the number that tells you least. A reel seen by four lakh people that sold nothing is a worse outcome than one seen by twelve thousand that sold four hundred units.
Decide the question before the campaign, not after
There are only three honest objectives, and they are measured differently:
- Awareness. You want people who did not know the brand to know it. Measure reach against new audience, plus branded search volume in the two weeks after.
- Consideration. You want people who know the brand to take it seriously. Measure saves, shares, comment sentiment, and traffic that lands on product pages rather than the homepage.
- Sale. You want units moved. Measure orders against a code or link, and accept that the code undercounts.
Picking all three means you will claim whichever one looks best afterwards. Pick one.
The metrics worth collecting
| Metric | What it tells you | Where it misleads |
|---|---|---|
| Reach / impressions | Scale of delivery | Says nothing about fit |
| Engagement rate | Whether the content landed | Inflated by giveaways and comment pods |
| Saves and shares | Genuine intent | Low absolute numbers, needs a baseline |
| Link clicks | Interest strong enough to leave the app | Lost entirely on in-app checkout |
| Code redemptions | Sales you can prove | Undercounts badly — see below |
| Branded search lift | Awareness that stuck | Needs a clean before-period |
Why your discount code undercounts
A code captures the people who remembered it at checkout. It misses everyone who saw the reel on Tuesday, searched the brand on Thursday and bought on Sunday without the code — which, for considered purchases, is most of them.
Two cheap corrections. First, run a holdout: keep one comparable city or audience out of the campaign and compare total sales, not coded sales. Second, add one question to your post-purchase flow — "where did you first hear about us?" — and treat it as directional, not exact.
A baseline makes the whole thing possible
Record fourteen days of normal before the campaign starts: daily sessions, daily orders, branded search, follower count. Without it every number afterwards is a number with nothing to compare to, and you will end up arguing about whether 3.1% engagement is good. It is good or bad only against your own last campaign.
What to do with a creator who underperforms
Look at the content before you blame the audience. A flat reel on a well-matched account is a briefing problem; strong engagement with no clicks is usually a landing-page problem; good clicks and no orders is a price or product problem. Only when the audience itself was wrong is the creator the wrong choice — and that is a selection question, covered in our guide to choosing nano, micro or macro creators.
The short version
Set one objective. Record a baseline. Report the metric that matches the objective, name the one that undercounts, and compare against your own previous campaign rather than an industry benchmark that was measured differently.