In short
- Fake or inactive followers show up as reach and views far below follower count, audiences in countries the creator never addresses, growth in sudden jumps and generic comments.
- Ask every creator for dated insights screenshots (audience location, age and gender, 30-day reach, Reel views) and cross-check them against what is public.
- No single signal proves fraud. Look for three or more together, and compare the creator with their own history and with peers of similar size.
- Protect the budget in the contract: an authentic-audience warranty, a right to verify, and payment after verified delivery.
Fake followers are accounts that follow a creator without being real, active people interested in the content: bots, click-farm accounts, abandoned profiles and follow-for-follow accounts. You spot them through patterns rather than one test: reach and Reel views far below the follower count, an audience in countries the creator never addresses, follower growth in sudden steps, and comments that could sit under any post. Confirm with the creator's own insights, never with the public profile alone.
This matters in India because brands are moving to judge creators on quality. In a 2025 Kantar-backed study reported by Campaign India, 85% of brands said they are shifting from follower counts to content quality. You cannot do that without checking whether the audience is real. This guide covers the types of inflated numbers, a table of red flags, the seven-step audit we use, the screenshots to request, tool limits and the contract terms that protect you. It sits alongside our complete guide to influencer marketing in India.
One honesty note: we do not quote a share of Indian followers that are fake. We found no reliable published India figure, and tool estimates are modelled. Anything claiming a precise national percentage deserves doubt.
Five ways numbers get inflated
"Fake" covers several different problems. Each one distorts different numbers, so identify which you are seeing.
| Type | What it is | What it does to the numbers | How it tends to look |
|---|---|---|---|
| Bought followers | Bots or click-farm accounts sold in bundles | Inflates followers; reach and views stay flat | Sudden jumps, empty profiles, odd names, wrong countries |
| Inactive followers | Real accounts that stopped using the app | Lowers engagement rate and reach per follower | Steady old growth, low reach, otherwise normal comments |
| Follow-for-follow and mass-follow | Accounts following thousands to win follow-backs | Inflates followers with people who do not care | High following count, mismatched audiences |
| Giveaway followers | Real people who followed for a prize | Spikes followers; engagement falls afterwards | Jump after a contest, comments like "done" or tagged friends |
| Pods and bought engagement | Groups or services that like and comment on cue | Inflates likes and comments, not reach or sales | Fast, generic comments from the same accounts |
Inactive and giveaway followers are real people, so the creator may not have done anything wrong. They still make the numbers misleading, so price the deal on verified reach and views and not on follower count.
Twelve warning signs
Group the signs by where you find them.
Followers
- Sudden follower jumps with no viral Reel, collaboration or feature to explain them.
- Audience mismatch. Large shares of followers from countries or cities the creator never talks to, or an age and gender mix that does not match the content.
- Hollow profiles. Many followers with no profile photo, no posts and random usernames.
- Following nearly as many accounts as followers, which suggests follow-for-follow.
Reach and views
- Reach or Reel views far below followers. A creator with 90,000 followers whose Reels land near 14,000 views has either an inactive audience or an inflated one.
- Story views far below followers, and flat from one story to the next.
- Views that spike on one post and sit low on all the others. One viral Reel can hide a weak average.
Engagement
- Likes with almost no comments, saves or shares.
- Generic comments ("nice 🔥", "great post", "follow me") or comments in scripts and languages the creator's audience does not use.
- Comments within the first minutes from the same accounts, post after post.
- Likes higher than views on a Reel, which cannot happen naturally.
Performance
- Brand posts that perform far below organic ones, or engagement that looks the same on every post whatever the content.
The red flag table
| Signal | What you see | Why it matters | How to verify | Weight |
|---|---|---|---|---|
| Growth spike | Follower chart steps up overnight | Often bought followers or a contest | Growth chart in insights; ask what happened that week | Medium |
| Wrong geography | Top cities or countries outside the creator's market | Followers who cannot buy from you | Audience location screenshot | Strong |
| Low reach per follower | Reach is a small fraction of followers on most posts | Inactive or fake audience | Reach for the last 10 posts in insights | Strong |
| Likes above views | Reel likes exceed plays | Bought likes | Compare likes with plays on the Reel | Strong |
| Generic comments | Emojis and praise only | Pods or bots | Read a 40-comment sample | Medium |
| Same commenters | Same 10 to 20 accounts on every post | Pod | Scan the first comments across five posts | Medium |
| Likes without saves or comments | High likes, near-zero saves and comments | Passive or bought engagement | Insights: saves and shares per post | Medium |
| Following count close to followers | A near-even ratio | Follow-for-follow | Public profile | Weak |
| Brand posts underperform | Sponsored posts far below organic | Audience not interested in the brand | Compare last five sponsored with last five organic | Medium |
| Edited screenshots | Odd fonts, missing dates, cropped numbers | Dishonesty | Ask for a live screen-share | Strong |
A single weak signal means little. Two medium signals together warrant a closer look. One strong signal, or three signals of any weight, is grounds to pause the deal.
The seven-step audit
Run these in order and stop when a creator fails clearly. The first two steps take about ten minutes and filter out most weak candidates before you spend any more time.
Step 1: the ten-minute public check
Open the profile and the last 12 posts. Note followers, following, average likes and comments, Reel views and whether the content fits the audience you want. Compute engagement with the engagement rate calculator and compare it with the tier ranges for India. Check views per follower on the last 10 Reels. Any creator far outside the range, high or low, goes on the "audit harder" list.
Step 2: check the growth history
Look at how followers grew over the last 90 days, from the creator's insights or an analytics tool. Natural growth is gradual with bumps tied to viral posts. A steep stair-step with flat sections between is the signature of bought batches or contests. Ask the creator what happened on the date of any jump; a real answer ("this Reel went viral", "we ran a giveaway") is normal, evasion is not.
Step 3: audit a sample of comments
Open five recent posts and read 40 comments in total. Sort each into a category.
| Category | Example | Healthy sign |
|---|---|---|
| Specific | References something in the content | Common |
| Question or intent | "Where can I buy this?", "Price?" | Present, especially on product posts |
| Generic praise | "Nice", "🔥🔥", "Awesome" | Some is normal |
| Promotional or spam | "Follow me", "DM for collab" | Rare |
| Off-language or off-script | Comments in languages the audience does not use | Rare |
| Repeat commenters | Same accounts on every post | A few loyal fans, not a bloc |
As a working rule, not a published threshold: if generic and spam comments make up most of the sample and specific or question comments are rare, treat it as a medium-to-strong signal. Real audiences also write long, messy, specific comments, often in Hinglish or a regional language, and they argue with each other.
Step 4: request insights screenshots
This is the step that separates guesses from evidence. Ask for the screenshots listed in the next section, dated, within the last 30 days. Compare them with public numbers: does reach match what the public posts suggest? Does the audience location match the language of the captions? Do the screenshots show the same account?
Step 5: compare with peers
Pull five to ten creators of similar size, niche and language, and compare views per follower, engagement by reach and comment quality. A creator that sits far below peers on views but above on likes is the pattern of bought engagement. Peer comparison also guards against false alarms: some niches simply have lower engagement.
Step 6: run a small paid trial
For a creator you like but cannot fully verify, book a single Reel with a unique link and code before any larger commitment. Track clicks, sessions and orders, not only views. If clicks arrive but sessions on your site do not, or if code redemptions cluster in a few minutes, the traffic may not be human. Our guide to measuring influencer marketing ROI covers the tracking setup.
Step 7: lock the protections into the contract
Write the audit into the agreement: the creator warrants the audience data is accurate, you may verify it, and payment depends on delivery. The clauses are below.
What to ask creators for
Ask politely and specifically, and explain it is a standard step for every creator. Most real creators have the data ready; some will send a media kit that already includes it. These are the screens to request on the professional dashboard. Menu names and layouts change with app updates, so describe the content rather than the button.
| What to request | Why |
|---|---|
| Accounts reached, last 30 days, with the follower and non-follower split | Shows how many people each post really reaches, and whether content travels beyond followers |
| Audience location: top cities and top countries | Confirms the audience can buy from you |
| Audience age range and gender | Confirms buyer fit |
| Followers growth chart, last 90 days | Reveals spikes |
| Reels plays and reach for the last 10 Reels | Calculates views per follower and the median |
| Saves, shares and comments for the same Reels | Shows quality of engagement |
| Story views for the last 10 Stories | A second view of audience attention |
| Past brand posts, with the brand's name and date | Lets you compare sponsored with organic |
Request screenshots with the date range visible, ideally the same day, and ask the creator to screen-share the insights on a short call if any number looks odd. Insights are available on professional accounts, so a creator who cannot show them is itself a signal, even if not a conclusive one.
Engagement pods and bought engagement, in detail
Pods are the harder problem, because they affect the metric brands trust most. Members agree to like and comment on each other's new posts within a short window, often through WhatsApp or Telegram groups. Bought engagement is a paid version of the same thing, using bots or low-paid workers.
Both leave marks you can see:
- Comments arrive in the first minutes after posting, then nothing for hours.
- The same 10 to 20 accounts appear under every post.
- Comments are generic or emoji-only and do not match the content.
- Likes are high but saves, shares and reach are low.
- Engagement is similar on every post, regardless of how good the content is.
The cost to you is more than wasted fees. Pod engagement can inflate engagement metrics without bringing buyers, and engagement from accounts that are not real or interested does not turn into sales. A creator whose real audience sees the post and ignores it will not produce the result you pay for, whatever the engagement rate says.
Tools: what they can and cannot tell you
Audience audit tools estimate the share of suspicious or inactive followers from public data. They are useful for ranking a long list and weak at proving anything about one account. Their models can flag real followers who happen to look inactive, and miss sophisticated fakes. Use one or two as a first screen, never as the verdict, and always confirm with the creator's insights. We do not rank tools: we found no independent test of their accuracy that we could cite.
How this fits the Gully Score
We score every creator out of 100 before a brand sees their name: audience authenticity 30, engagement quality 25, buyer fit 20, content craft 15 and brand safety 10. The method is on our how we work page. Authenticity and engagement together carry 55 points, which is deliberate. A creator needs at least 70 to be shortlisted, and the other three criteria add up to only 45, so even a perfect fit needs at least 25 of those 55 points to reach 70. A creator who fails the audience check cannot score their way onto a shortlist with strong content alone.
Contract protections
| Clause | What it says | Why it helps |
|---|---|---|
| Audience warranty | The creator confirms insights data is accurate and unedited | Gives you a remedy if screenshots were doctored |
| Right to verify | You can request live insights during the campaign | Lets you check delivery |
| Metric definition | Views, reach and the date range are defined | Avoids arguments over numbers |
| Payment on verified delivery | Part of the fee is paid after delivery is confirmed | Limits your exposure |
| Replacement or refund | Remedy if delivery is materially below the quoted figures | Keeps the budget intact |
| Takedown and non-disclosure remedies | Rights if a post is removed or poorly labelled | Protects brand safety |
Our guide to influencer contracts in India lists the full set of clauses.
What to do when you find a problem
| Finding | Likely cause | Action |
|---|---|---|
| Low reach, otherwise healthy comments | Many inactive followers | Re-price on verified views and keep the creator |
| Generic comments and repeat commenters | Pod | Drop, or retest with a small paid trial |
| Growth spike, audience from unrelated countries | Bought followers | Drop; note it for future shortlists |
| Edited or inconsistent screenshots | Misrepresentation | Drop and stop discussions |
| Clicks with no sessions, or redemptions in minutes | Bot traffic or code abuse | Pause the creator and audit tracking |
Keep a record of why you dropped each creator, so you do not re-audit the same account next quarter. If you have already paid, use the contract's verification and refund terms.
Common mistakes
- Relying on a public engagement rate alone. It can be faked, and it hides reach.
- Judging by one tool score. Tools estimate.
- Auditing only large accounts. Numbers can be inflated at any size, so audit every creator. Nano and micro creators are the largest group: 61.1% of creators surveyed in a 2026 vendor study covered by Campaign India were nano.
- Assuming a real follower is a buyer. Check where followers live and whether they speak the language of your campaign.
- Skipping the contract terms. Without them, a failed audit after payment is an argument, not a refund.
If you want this audit done for you, our micro-influencer marketing service vets every creator before you see a name, and you can send us a brief to get a vetted creator plan within 48 hours.
Sources
- Indian consumers and influencers: coverage of a 2025 Kantar-backed study of consumers and brands, Campaign India (12 Jun 2025)
- Coverage of a 2026 India creator-economy report: creator counts by tier, Campaign India (21 May 2026)
- Micro influencer rates in India 2026: what to charge (average views rule of thumb), Boss Wallah (14 Aug 2026)
Frequently asked questions
How can I tell if an influencer has fake followers?
Compare reach and Reel views with follower count, read 40 comments for generic or off-topic replies, check follower growth for sudden jumps and ask for insights screenshots showing audience location and age. One signal is a hint; three or more together are a strong case. Our engagement rate calculator covers one part of the check.
What is a fake follower?
It is an account that follows a creator but is not a real, active person interested in the content: a bot, an account from a click farm, an abandoned profile or a follow-for-follow account. Real but inactive followers and giveaway followers are different. They are real people who rarely engage, and they distort numbers in a similar way.
What are engagement pods?
Pods are groups of creators, often organised in WhatsApp or Telegram groups, who agree to like and comment on each other's posts to boost engagement. They produce quick, generic comments from the same accounts and little real reach or purchase intent. Pod engagement raises the engagement rate without raising sales.
Are fake follower checker tools accurate?
They give estimates based on public data and models, so treat them as a screen, not proof. False positives and false negatives both happen. Use them to rank a long list, then verify the shortlist with the creator's own insights and a comment review.
What should I ask an influencer to prove their audience is real?
Ask for dated screenshots of audience location, age and gender, accounts reached over the last 30 days with the follower and non-follower split, and Reel plays for recent posts. For anything doubtful, ask them to screen-share the insights on a call. Our how to find influencers guide has the full checklist.
What percentage of followers is fake in India?
We do not quote one. We found no reliable published India figure, and tool estimates are modelled and vary widely. Rather than relying on a platform-wide average, audit each creator against their own reach, comments and audience data, and compare them with peers of similar size and niche.
What do I do if I find a creator has fake followers?
If the audience is mostly inactive but the creator is otherwise good, re-price the deal on verified views. If you find pods, retest with a smaller paid trial. If the creator supplied edited screenshots or bought followers, drop them. Contract terms such as a right to verify and payment on delivery limit your loss. See our guide to influencer contracts in India.