Instagram follow-back statistics (2026): data from 4,609 accounts

Key numbers

Breakdown by account size

Accounts you follow that don't follow you back

22% of accounts have half or more of their following list not following back (32% of accounts with under 300 followers). 42% have 100 or more accounts that don't follow back (61% of accounts with 1,000 to 2,999 followers).

Following more people than follow you

Weeks that ended with fewer followers

Secondary finding: followers you don't follow back

It works both ways. The typical account doesn't follow back 39% of its own followers (median; middle half of accounts 20% to 62%). That is more than the share of its following list that doesn't follow it back (21%). Pooled across all accounts, 51% of follower relationships are not returned.

• Under 300 followers: 27%
• 300 to 999 followers: 38%
• 1,000 to 2,999 followers: 55%
• 3,000 and up followers: 79%

Definition: followers not in the following list, divided by the follower count. Sample: 4,609 accounts. Time window: lists collected December 16, 2025 to September 5, 2026.

Context: who is in these lists

Part of the reason so many follows aren't returned: about 1 in 5 accounts people follow carries a blue check. In a typical following-list sample, a median 21% of accounts are verified (middle half 9% to 39%), against 2% in a follower-list sample (965 and 935 accounts). Verified now includes Meta Verified subscribers.

Private accounts are common among followers: in a typical account's follower sample, a median 30% are private (middle half 18% to 43%; 935 accounts, about 122,000 profiles). In following lists the median is 10%.

Time window: September 10 to October 9, 2026. Both numbers come from the first page of each list we read (about 45 profiles), which is not a random sample of the list, so treat them as context rather than headline figures. Private status is as of the moment it was read.

What we don't publish yet

We don't publish unfollow rates, how soon people unfollow after following, or which days people unfollow. Measuring those honestly needs repeated complete follower lists for the same accounts over several weeks, and this study doesn't have enough of them yet. We'll add those numbers when it does.

How we got these numbers

• Source. Accounts that signed-up users entered into Unfollowers Track, using public profile data from our own scans. The numbers were pulled on October 9, 2026 with read-only database queries.
• Aggregates only. Only counts, shares and percentiles left the database. No usernames, IDs, handles or profile links appear anywhere in the results, and we never report the largest or smallest single account.
• At least 50 accounts per group. Every reported group has k ≥ 50 accounts. Smaller groups are suppressed or merged.
• Each account counts once. We use the latest profile, the latest complete list and the latest reading per account. The weekly number counts account-weeks (up to 4 per account).
• Completeness filter. Follow-back numbers only use accounts whose last follower and following lists both finished and returned 98% to 102% of the count Instagram showed. 4,609 of 4,949 candidate accounts pass. When the profile was refreshed the same day, a list that passes matches Instagram's follower count within 2% in 99.9% of cases (3,749 accounts). A stricter filter (exact count match, 2,094 accounts) gives the same picture.
• Follower-count changes. These use the follower count Instagram displays, checked about every 6 hours, not list comparisons. 6-hour changes larger than 20% of followers are treated as glitches and dropped. About 11% of 6-hour drops are exactly undone in the next reading, so the "strict" figures ignore those.
• Time windows. Follower and following lists: December 16, 2025 to September 5, 2026. Follower-count checks: September 10 to October 9, 2026. Profile counts: refreshed January 2 to October 9, 2026.
• Rounding. Shares are rounded for reading. 95% confidence intervals are given where they matter.

Limits of this data

1. Self-selected sample. These are accounts people typed into an unfollower tracker: their own, a partner's, an ex's, a friend's. They are not a random sample of Instagram.
2. Small, public accounts. Private accounts can't be scanned, and at the time of this data our scans were capped at 10,000 followers or following (3,000 before June 2026). The median account has about 630 followers. These results say nothing about creators or brands.
3. Geography skews to Southeast Asia. Most of our site's visitors are in Southeast Asia. We don't store a country for scanned accounts, so we can't break the numbers down by country.
4. "Doesn't follow back" includes celebrities and brands. About 21% of a typical following-list sample is verified accounts, and those rarely follow back. That is real, but it isn't friends ignoring you.
5. Net change is not unfollows. Instagram's follower count moves by new followers minus lost followers. A net drop proves at least one follower was lost, so the count-change numbers are lower bounds. A lost follower can also be an account that was deactivated, deleted or removed by Instagram.
6. Change numbers cover monitored subscribers only. The weekly and 3-day numbers come from accounts under paid or trial monitoring, checked every 6 hours. Many stop after a short trial, so these samples are small.
7. Snapshot timing. Follow-back numbers are measured at each account's last complete list check (December 2025 to September 2026), not on a single day.