How to calculate Instagram follower growth

Your follower count on its own does not tell you much. Growth — how fast that count is changing — does. Here is how to calculate it properly and compare periods fairly.

The basic formula

growth rate = (followers at end − followers at start) ÷ followers at start × 100

Example: you had 4,000 followers on 1 March and 4,320 on 31 March. Growth = (4,320 − 4,000) ÷ 4,000 = 0.08, or 8% for the month.

Comparing periods of different lengths

Eight percent in 30 days is not the same as eight percent in 90 days. Two ways to make periods comparable:

• Average daily change: net change ÷ days. In the example, 320 ÷ 30 ≈ 10.7 followers per day.
• Monthly equivalent rate: convert the period's growth into the rate it implies over 30 days, using compounding: (end ÷ start)^(30 ÷ days) − 1.

Why compounding?

Growth usually scales with size: more followers means more people who can share and recommend your posts. Compounding captures that. For short periods and small rates the difference from simple division is tiny; over longer horizons it matters.

Net growth is two numbers in disguise

Your count changes by new followers minus lost followers. Measuring them separately is far more informative:

• Gross growth = new followers ÷ starting followers — how well you attract people.
• Churn = lost followers ÷ starting followers — how well you keep them.

Two accounts with the same net growth can have very different health. Use the gain and loss calculator to split them.

A simple monthly routine

1. Record your follower count on the same day each month.
2. Calculate the month's growth rate.
3. Note anything unusual: a viral post, a break from posting, a giveaway.
4. Look at the trend over three to six months rather than any single month.

The follower growth calculator does the arithmetic, including projections — and UnfollowWatch records your follower count automatically each time you scan.

Common mistakes

• Comparing a week with a month without normalising.
• Reading a sudden platform-wide clean-up as a content problem.
• Treating projections as forecasts — they assume nothing changes.