How to Vet Influencers Before You Pay: Spotting Fake Followers and Bought Engagement

Follower counts are cheap to inflate. Followers, likes, comments and even views can be bought in bulk, and accounts that have been pumped up this way can look impressive at a glance. If you pay a creator based on numbers that are partly fake, you are not just wasting the fee; you are also drawing wrong conclusions about what works for your brand, because the campaign "reached" people who do not exist.

The good news is that most inflated accounts give themselves away if you spend twenty minutes looking properly. This guide is a manual process you can run on any Instagram or YouTube creator before you agree a fee. It does not depend on paid analytics tools, and it works just as well for a nano creator with 4,000 followers in Coimbatore as for a lifestyle creator with 2 lakh followers in Mumbai.

Why manual checks still matter

Third-party tools that estimate "fake follower percentages" can be a useful signal, but they work from public data and their own models, and they can be wrong in both directions. They also cannot see the numbers that matter most: who actually watched a Reel, from which cities, and for how long. Only the creator can see that, inside their own Insights or YouTube Studio.

Many of the ways brands find creators, such as Instagram DMs, agency spreadsheets and marketplaces that do not connect to social platforms' APIs, give you no automatic access to that audience data at all. The responsibility for checking sits with you, and the method is a combination of looking at public signals yourself and asking the creator for evidence.

Step 1: Look at the profile like a stranger would

Before you open any numbers, scroll the profile for a few minutes and ask simple questions:

• Does the bio clearly say who this person is and what they post about?
• Is the content consistent in topic and quality over the last few months, or does it jump around?
• Are there real faces, places and voices, or mostly reposted memes and stock images?
• How often do they post? Long gaps followed by bursts can be normal, but note them.
• How many recent posts are sponsored? A feed where almost every post is an ad tends to have an audience that has learnt to ignore ads.

This step will not catch sophisticated fakes, but it quickly filters out accounts that are clearly not a fit for your brand, which saves time on the deeper checks.

Step 2: Compare views to followers

On short-form video, views are harder to fake convincingly at scale than followers, and they tell you how many people actually see a creator's content today. Open the Reels tab (or YouTube Shorts tab) and note the view counts on the last 10–15 videos.

• Look at the typical video, not the best one. One viral Reel from eight months ago tells you little about what your post will do. Use the median of recent videos.
• Compare that median to the follower count. There is no universal "correct" ratio, because it varies by niche, format and how the algorithm is treating the account. But if an account has 1.5 lakh followers and most recent Reels get a few thousand views, a large share of those followers either are not real or are no longer paying attention. For your purposes, both mean the same thing.
• Watch for suspicious consistency. Real view counts are uneven. A long run of Reels that all sit at almost exactly the same view count, regardless of topic or quality, can be a sign of purchased views.

Price creators on realistic views rather than followers. A smaller account that reliably gets watched is usually better value than a large account that does not.

Step 3: Check engagement, then read the comments

Calculate a rough engagement rate on recent posts: likes plus comments, divided by followers (or, better, by views for video). Our free engagement rate calculator does the arithmetic for you. Very low engagement is a warning sign, but so is engagement that looks unnaturally high for the account's view counts, because likes and comments can be bought too.

The number matters less than the quality of the engagement. Open the comments on five or six recent posts and read them properly.

Signs of genuine engagement

• Comments that reference something specific in the video ("which shade is that lipstick?", "is this café open on Mondays?").
• Questions, and the creator replying to them.
• Conversations between followers, in-jokes, regulars who comment often.
• Comments in the language and slang you would expect from the creator's audience, including regional languages and Hinglish.

Signs of bought or pod engagement

• Lots of generic comments: "Nice", "Wow", "Great content", strings of fire or heart emojis.
• Comments that do not match the post ("Beautiful picture!" on a talking-head video).
• The same handles commenting first on every post within minutes, often other creators in an engagement pod.
• Commenters whose own profiles have no posts, no profile picture, random-letter usernames, or are clearly from a region that makes no sense for the content.
• A sudden spike of likes in the first minutes and then almost nothing.

Engagement pods (groups of creators who agree to like and comment on each other's posts) are not exactly fraud, but they inflate engagement with people who are never going to buy your product.

Step 4: Sample the followers

Open the follower list and scroll through a sample of a hundred or so accounts. You are not looking for perfection; every account has some bots. You are looking for patterns:

• A large share of accounts with no profile photo, zero posts and auto-generated usernames.
• Clusters of followers from countries that have nothing to do with the creator's content or language.
• Lots of business or "follow-for-follow" accounts in an unrelated niche.

If you can, also look for signs of sudden follower jumps. Some public tracking sites show historical follower counts; a smooth upward curve is normal, while a jump of tens of thousands in a few days without a viral post to explain it is a common sign of purchased followers. Treat these sites as a rough indication, not proof.

Step 5: Ask for Insights, the right way

Public signals only get you so far. The most important data, audience location, age, gender, reach and watch time, is only visible to the creator. Ask for it. Professional creators expect this request and will not be offended; reluctance to share is itself useful information.

What to ask for

1. Account-level audience: top cities, top countries, age range and gender split.
2. Account reach over the last 30 days, and the split between followers and non-followers.
3. Post-level insights for two or three recent Reels and one recent sponsored post: plays or views, reach, average watch time, saves, shares and, if there was a link, link taps.
4. For YouTube: YouTube Studio analytics for views, average view duration, audience geography and age for the last 28 or 90 days.

Screenshots vs. screen recordings

Screenshots are easy to share and easy to edit. For larger fees, ask for a short screen recording instead: the creator opens the app, navigates from their profile to Insights and scrolls through the relevant screens in one continuous clip. It takes them a minute and is much harder to fake. A live video call where they share their screen works too.

When you receive screenshots, check them:

• Do the handle and profile picture match the account you are vetting?
• Is the date range visible and recent?
• Do the numbers roughly agree with what you see publicly? If a Reel shows 20,000 views publicly, the Insights should not claim 2 lakh plays.
• Are fonts, alignment and spacing consistent with the current app interface? Edited figures sometimes sit slightly off.

Step 6: Check audience fit, not just authenticity

An account can be completely genuine and still be wrong for you. Once you have Insights, compare the audience to the customer you want:

Step 7: Review past brand work

Look at the creator's last few sponsored posts. Were they disclosed clearly? Did the creator seem to actually use the product, or did it feel like a read-out? How did the audience react in the comments, with genuine questions or with complaints about too many ads? A creator who handles sponsored content well will make your campaign look natural; one who does not will make it look like an advert their followers scroll past.

Also check disclosure habits. A creator who routinely hides paid partnerships creates risk for the brands they work with. For the current Indian guidelines on influencer disclosure, refer to the Advertising Standards Council of India at ascionline.in.

Red flags that should stop or slow a deal

• Refusing to share any Insights, or only sending cropped screenshots without handle or dates.
• Insights numbers that contradict public view counts.
• Follower count far ahead of typical views, with no explanation (such as a long break from posting).
• Comment sections dominated by generic or irrelevant comments.
• Large, unexplained follower jumps.
• Insisting on full payment upfront for a first collaboration, with no deliverables agreed in writing.
• Pressure tactics: "the rate goes up tomorrow", "another brand wants this slot".
• Unwillingness to label the post as paid or gifted.

One red flag on its own is not always disqualifying; there are innocent explanations for many of them. Ask about it. A genuine creator will usually have a straightforward answer.

Reduce risk with how you structure the deal

Vetting lowers the risk; deal structure limits the damage if you still get it wrong.

• Start small. Book a single Reel or a Story set before committing to a multi-post package.
• Agree reporting upfront. Make post-campaign Insights screenshots a deliverable, due a fixed number of days after posting.
• Split payment. A part on acceptance and the balance after posting is fair to both sides. Our guide to creator–brand payment terms and contracts covers common structures.
• Use tracked links and codes so you can see real outcomes, not just reported reach.
• Prefer more, smaller creators early on. Spreading spend across several nano and micro creators means one bad pick does not sink the campaign. Our comparison of micro vs. macro influencer strategies goes into the trade-offs.

A vetting checklist you can reuse

Copy this into a spreadsheet with one row per creator and one column per check:

1. Profile and content are consistent and relevant to our category.
2. Median views on last 10–15 videos noted; reasonable relative to followers.
3. No suspiciously uniform view counts across videos.
4. Engagement rate calculated on recent posts.
5. Comments read on 5+ posts; mostly specific and genuine.
6. Follower sample checked; no large clusters of empty or irrelevant accounts.
7. No large unexplained follower jumps.
8. Audience Insights received (screen recording for larger fees).
9. Insights match public numbers, handle and recent dates.
10. Top cities, age range and language fit our customer.
11. Past sponsored posts reviewed; disclosed clearly; audience reaction positive.
12. Deliverables, reporting and payment schedule agreed in writing.

Over time this sheet becomes valuable in its own right: you will see which checks actually predicted good results for your brand and which you can safely skip.

Where to find creators to vet

Vetting is easier when you have several options to compare instead of one creator you have already fallen for. On Kollabary, brands post a campaign and creators bid on it with their own proposals, so you can review a shortlist side by side and ask each one for Insights through direct messages before agreeing anything. Kollabary does not pull data from Instagram or YouTube, and while creator verification is done manually, it is a starting point, not a substitute for the checks above. Fees are agreed and paid directly between you and the creator. If you have not written your brief yet, start with our guide on writing an influencer campaign brief.

Key takeaways

• Follower counts are the easiest number to fake; judge creators on recent median views instead.
• Read comments, not just counts. Specific questions and conversations signal a real audience.
• Always ask for Insights; for bigger fees, ask for a screen recording rather than screenshots.
• Check that numbers in Insights agree with what you can see publicly.
• A genuine audience in the wrong city, age group or language is still the wrong audience.
• Start with small, tracked deals and split payments to limit risk while you learn.