To check an Instagram account for fake followers, open the Potential spam list Instagram keeps inside your follower list, then compare follower growth, top follower locations and how much of your engagement comes from followers in Insights, looking for spikes that don't match anything you posted. Clear obvious spam with Instagram's own Remove all spam followers option, keep the spam filter on and leave quiet but real followers alone. If your count suddenly falls, Instagram says it may simply have removed inauthentic activity or disabled accounts.
This guide is about auditing the followers you already have. Buying followers or engagement, and the US rule against it, is covered in our guide to fake engagement rules, and the arithmetic of engagement rates lives in our Instagram engagement rate worksheet.
Why follower counts drop on their own
Instagram's article on changes to your likes, follows or comments gives two reasons your numbers can move without you doing anything. Non-Instagram apps sometimes generate automated likes, comments and follows to make accounts look more popular, which breaks Instagram's guidelines, and when those accounts are discovered their activity is removed. Instagram also routinely removes disabled accounts, so your follower and following totals can change, sometimes within a short time.
A drop of that kind takes away nothing that was real. It deserves a closer look in two situations: if you once used an app that promised followers or likes, or if the drop arrives alongside blocked actions. Instagram's third-party apps page says engagement generated by those apps may be removed and that accounts which keep using them may find parts of the account limited. Our guide to Instagram account restrictions covers what to do in that case.
Instagram's own spam-follower tools
Instagram filters likely spam into a separate Potential spam list inside your followers, and its help page on spam and bot followers says accounts that are likely to be spam can't follow you without your confirmation. Go to your profile, open your follower list and tap Potential spam. You can confirm a request, delete it, or tap Remove all spam followers to clear the list in one pass, and Instagram says it won't tell those accounts they were removed.
The filter lives in Settings and privacy, under Followers, as a toggle called Filter spam followers. Leave it on unless you want to vet every follow by hand. It only catches what Instagram's systems flag, which is why the worksheet below adds a manual sample of your newest followers.
The follower-quality audit worksheet
Run the worksheet each quarter, or after any growth spike you can't explain. It tells you what to compare rather than what number to hit, so the notes column is yours to fill in. Instagram's About Instagram insights page says follower trends, including growth, top locations, age range and most active times, appear once you have at least 100 followers, and that you can view any period within the past 90 days.
| Check | Where to find it | A normal pattern | Worth a closer look | Your notes |
|---|---|---|---|---|
| Growth spikes | Insights, Total followers, then the growth chart for your chosen range | Gains that line up with a reel, collab or mention you can name | A jump on a day you posted nothing, or follows that arrive and then vanish | Dates of each spike and the post you think caused it |
| Top locations | Insights, Total followers, top locations of your followers | Countries and cities that fit your language, content and timezone | A country you never post for sitting near the top of the list | The top places and whether each one makes sense |
| Follower share of engagement | Post, story and Live insights, where Accounts engaged splits into followers and non-followers | Followers engaging with a fairly steady share of what you post | The follower count climbing while follower engagement stays flat | Follower and non-follower split on recent posts |
| Potential spam list | Profile, follower list, Potential spam | A short list you clear now and then | The list filling up again soon after every clear-out | Date cleared and roughly how full it was |
| Manual sample | Your follower list, newest followers first | Recent followers with photos, posts and signs of real activity | Clusters of blank profiles, machine-looking usernames or identical bios | How many of your sample looked automated |
| Comment quality | Comments on your most recent posts and reels | Replies that refer to what the post actually shows or says | Generic praise, emoji strings or link spam from strangers | Any repeat offenders to remove or block |
If you want the follower and non-follower split explained in more depth before you start, our guide to Instagram Insights for creators walks through each screen.
Signs a follower probably isn't a person
No single sign proves an account is fake, so look for several at once. Meta's inauthentic behaviour policy describes the deception it removes as accounts used to mislead people about identity, origin or the popularity of content, and these are the traces that behaviour tends to leave:
- No profile photo, no posts and no highlights, combined with a very large following list.
- A username that looks machine-made, such as a first name followed by a long run of digits.
- A bio that is empty, copied word for word from other new followers, or nothing but a link.
- A DM or comment pushing a link, a giveaway or an investment offer soon after the follow.
- A wave of near-identical profiles following within the same hour.
- An account that follows, unfollows and follows again to get your attention.
Real people can tick one or two of these boxes. Someone who never posts may still watch every reel and recommend you to friends, so treat the list as a reason to look closer rather than a rule for removal.
Remove or keep: a decision table
| Follower type | Decision | Reasoning |
|---|---|---|
| Accounts sitting in Potential spam | Remove all | Instagram has already flagged them, and the removal is silent |
| Accounts sending scam links or offers in DMs or comments | Remove, block and report | They put your audience at risk as well as distorting your figures |
| Blank profiles from a spike you can't explain | Remove in batches after checking a sample | They add to the count without adding reach, and they blur your Insights |
| Real people who rarely interact | Keep | Quiet followers still see and share your work; removing them gains nothing |
| Real followers from countries you didn't expect | Keep, then find out why | A reel may have travelled; check which post brought them in |
| Follow-for-follow and engagement-group accounts | Your call | They are real, but they followed for a return favour rather than your content, so they flatter your numbers |
| People who followed to enter a giveaway | Keep if real, and review how the giveaway worked | Meta's spam policy bans offering cash or anything of monetary value in exchange for engagement |
When a bot wave arrives that you didn't ask for
Sometimes a burst of empty accounts follows a creator who bought nothing, whether from a spam network testing accounts or someone trying to make you look bad. Instagram's filter is the first defence, since flagged accounts can't follow you without your confirmation. For anything that slips through, write down the date and what you saw, take a screenshot of the growth chart, then remove the accounts in batches. A dated note is useful if a partner later asks about a spike, because it shows you noticed it and acted rather than kept the inflated number.
Afterwards, a light routine keeps the base clean without turning it into a weekly chore:
- Clear the Potential spam list whenever you open your follower list for another reason.
- Check the growth chart after every reel that travels further than usual, while the cause is still obvious.
- Repeat the manual sample of newest followers before you send a media kit or pitch.
- Recalculate your engagement rate after a large clean-out, so the figure you quote reflects the audience you actually have.
Why audience quality matters to brand partners
A brand paying for reach is paying to be seen by real people, so a follower total padded with bots overstates what you are selling. Cleaning out obvious spam before you pitch keeps your follower count, engagement rate and audience locations consistent with each other, and makes any Insights screenshot you share easier to stand behind. Our influencer media kit template shows which figures to include and how to date them.
In the US, the Federal Trade Commission's rule on consumer reviews and testimonials defines fake indicators of social media influence to include followers generated by bots, accounts not tied to a real person and hijacked accounts, and it bars buying or selling them where they misrepresent influence for a commercial purpose. That rule is aimed at the trade in fake metrics, which our fake engagement guide covers; for an audit, its definition is a handy test of what to clear out.
Keep audit tools away from your login
Plenty of services offer to scan your followers for you. Instagram's third-party apps page warns against handing any app your username and password, because it can then gain complete access to the account, read your messages and post spam, and it says you shouldn't use apps that offer likes or followers at all. Instagram's own Potential spam list and Insights need no outside access. If you do try an outside tool, connect it through the official login rather than a password, and remove it under Apps and websites in Settings once you have what you need.
Limitations of this audit
Instagram doesn't publish how it decides an account is likely spam, and its Insights figures aren't exact: the insights help page says some metrics are estimated and still in development. A removal can only be reversed by that person choosing to follow you again, so an over-eager purge costs real followers.
The worksheet is a judgement aid built on Instagram's tools as they appeared on 1 October 2026, not a score. It can't detect paid engagement designed to look human, and it says nothing about whether a particular brand will accept your figures.