Fake followers and bought engagement are cheap and easy to add to any Instagram account, so a creator's follower count and engagement rate tell you very little on their own. Vet a creator on the signals that are hard to fake: the shape of their follower growth, where their audience actually lives, story views relative to followers, and whether comments respond to the content. Then ask for live insights, start with one paid deliverable, and release most of the payment only after the post goes live.
Why Engagement Rate Is the Easiest Number to Buy
Most D2C brands in India still shortlist creators the same way: a follower range, an engagement rate above some threshold, a quick scroll through the grid to check the aesthetic fits. The trouble is that the two numbers doing the filtering are the two easiest numbers in influencer marketing to manufacture. Followers can be bought in bulk. Likes and comments can be bought per post, or traded through engagement pods where groups of creators agree to like and comment on each other's content within minutes of it going live.
None of this is exotic. It is widely available, cheap relative to what a brand pays for a single reel, and the incentive to use it is obvious: brands and agencies screen on engagement rate, so engagement rate is what gets inflated. The result is a shortlist that rewards the creators most willing to game the metric, and a campaign that looks healthy in the report and does nothing for sales.
The fix is not a better threshold. It is screening on signals that cost a fraudster far more effort to fake than they cost you to check.
| What the media kit shows | What is much harder to fake |
|---|---|
| Follower count | Follower growth curve over 12 months |
| Average engagement rate | Audience cities, language and age |
| Screenshots of past reel views | Story views as a share of followers |
| A list of brands worked with | Comments that respond to the content |
What Bought Followers and Engagement Actually Look Like
No single signal proves fraud. What you are looking for is a pattern where several numbers that should move together do not.
Follower growth that jumps without a reason. Genuine accounts grow unevenly, usually in steps that line up with a reel that travelled or a collaboration with a bigger creator. Bought followers tend to arrive as sudden jumps with no post to explain them, sometimes followed by a slow decline as the platform removes inactive accounts. A growth history is available through most third-party audit tools, and it is the single most useful thing to look at before anything else.
Reach that does not match audience size. A creator with a large following whose stories are watched by a small fraction of it, or whose reels consistently underperform what the follower count suggests, is carrying a lot of followers who are not really there. Story views are particularly telling, because they are shown mostly to existing followers and are harder to inflate than likes.
Comments that could sit under any post. Read the comments on the last ten posts rather than counting them. Strings of fire emoji, "nice pic", "amazing content" and the same handful of accounts appearing within minutes of every upload point to pods or bought engagement. Real comments ask where the kurta is from, argue with the creator's opinion, tag a friend, or respond to something specific in the video.
An audience that lives somewhere else. A creator who makes Hindi content about Jaipur street food, with an audience concentrated in countries or cities that have no reason to watch it, has either bought followers or grown through giveaways that attracted the wrong people. Either way, those followers will not buy your product.
Asking for Proof Without Insulting the Creator
Once a creator clears the public screen, ask for first-party data. The request that works best is a live screen recording of their account insights, scrolling through audience cities, age and gender split, and reach and story views for the last thirty days. Screenshots are easy to edit and easy to pick selectively; a screen recording of the app is much harder to fake and takes the creator two minutes.
Framing matters. Present it as standard process applied to every creator on the campaign, which it should be, rather than as suspicion of this particular person. Most creators with a genuine audience are used to the request and share without hesitation, because good numbers help them negotiate. A creator who refuses, delays repeatedly or sends only cropped screenshots has told you something useful without you having to accuse anyone.
It is also worth asking what happened on one or two past brand collaborations: which post, roughly what it delivered in reach and saves, and whether the brand came back for more. Repeat collaborations with the same brands are one of the better signs of a creator who actually moves product, because brands rarely rebook someone who did nothing for them. This matters more at the smaller end, where what nano and micro creators in India actually charge is low enough that brands often skip vetting entirely and end up paying for dozens of unverified audiences at once.
"Screen on the numbers that are expensive to fake, not the ones that are expensive to check."
- Brand Integer Influencer Marketing Team
A Vetting Sequence Before Any Payment
Put together, vetting becomes a short sequence you run on every creator before any money moves. It adds a few days to a campaign timeline, which is far cheaper than a campaign built on audiences that do not exist.
- Screen the public profile Growth spikes, comment quality, views against followers
- Ask for live insights Screen-recorded audience and reach data, not screenshots
- Check audience fit Cities, language and age against your actual buyers
- Start with one paid test A single deliverable before any multi-post package
- Pay against delivery Milestones, post-live insights and a make-good clause
The last two steps carry most of the protection. A single paid deliverable before any multi-post package means the first collaboration tests the audience before you commit to the larger fee. Paying the bulk of the fee after the content goes live, with post-live insights as a condition of payment, means you see delivered reach before the money leaves rather than after. A make-good clause, committing the creator to an extra post or story if delivered reach falls well short of what their shared insights suggested, closes the gap for the cases where the numbers disappoint. These terms belong in writing alongside usage rights and exclusivity, and the payment and exclusivity terms that belong in an Indian influencer contract covers how to phrase them so they hold up.
For brands testing many small creators at once, a gifting round can act as a cheap first filter before any paid work, because the creators who post honestly about a product they were sent, and whose audiences respond, identify themselves. The trade-offs of that approach are covered in when product seeding works better than paid collaborations.
Inflated Is Not Always Fake, and Real Is Not Always Right
Not every creator with weak numbers bought them. Many built their audience through giveaways, follow-for-follow trends or a single viral post years ago, and are left with a following that is real but inactive. They are not committing fraud, but for a brand the outcome is the same: you are paying for reach you will not get. Judge the audience you can reach today, not the history of how it was assembled.
The opposite mistake is just as common. A creator can pass every fraud check, with clean growth, honest comments and a real audience, and still be the wrong choice because that audience lives in the wrong cities, speaks a different language from your packaging or cannot afford your price point. Fraud screening removes the fake audiences. Audience fit is what decides whether the real ones buy.
That is why vetting should connect directly to how the campaign is measured afterwards. If you have screened for audience fit, you know what the post should have delivered, and measuring influencer marketing ROI in India becomes a comparison against that expectation rather than a guess. Over a few campaigns, the creators who repeatedly deliver against their own shared insights become your core roster, and the vetting work gets lighter every quarter.
Frequently Asked Questions
How can I tell if an Indian influencer has fake followers?
Look at the shape of their follower growth rather than the total. Real accounts usually grow in uneven steps that line up with content that did well, while bought followers tend to arrive as sudden jumps with no post to explain them. Then compare reach with audience size: story views and reel views that are very low relative to follower count, comments that are generic emoji or unrelated praise, and an audience concentrated in cities or countries that do not match the creator's language and content are all signs worth questioning before you pay.
What should I ask an influencer for before a paid collaboration?
Ask for a live screen recording of their account insights rather than screenshots, covering audience cities, age and gender, and reach and story views over the last month. Ask for results from one or two past brand collaborations, ideally with the brand's permission to share them. A creator with a genuine audience will usually share this without difficulty, and hesitation or a refusal is useful information in itself.
Are influencer audit tools reliable for Indian creators?
They are useful as a first filter, not as a verdict. Most tools estimate audience quality from public signals, so they can flag obvious bought followers and suspicious growth, but they are weaker at judging audience fit, regional language audiences and whether engagement is organic but irrelevant to your product. Use a tool to shortlist, then ask the creator for first-party insights and check the comments yourself before committing budget.
How do we protect payment if a creator's numbers turn out to be inflated?
Structure payment so that most of it is released after the content goes live, and make sharing post-live insights a condition of that payment. Include a make-good clause that commits the creator to an additional post or story if delivered reach falls well short of what their shared insights suggested. Start with a single paid deliverable before agreeing to any multi-post package, so the first collaboration tests the audience before the larger commitment.
