X (Twitter) Audience Loyalty Calculator

The calculator needs two counts: how many different accounts engaged with your posts, and how many of them did so on 2 or more posts. Both come from a tally you make yourself.

1. Pick one period, such as the last 28 days, and a fixed set of posts inside it.
2. List every account that replied to, reposted, quoted or liked those posts. Count each handle once. That total is the first field.
3. Mark each handle that appears on 2 or more posts. That total is the second field.
4. Add followers, and the same tally for an earlier period, to unlock the other ratios.

Audience loyalty on X is how often the same people come back to your posts. X has no metric with that name and publishes no loyalty score. This page uses the term for ratios that count returning people.

The X developer documentation, read on October 3, 2026, lists the post metrics X reports: impressions, likes, replies, reposts, quotes and bookmarks. Each one counts actions. None of them says whether 100 likes came from 100 people once or from 20 people five times. Loyalty is about that difference, so it needs a count of people.

The ratios on this page use three kinds of people:

• Engagers: different accounts that engaged with at least one post in the period.
• Repeat engagers: engagers that showed up on 2 or more posts in the period.
• Returning engagers: accounts that engaged in an earlier period and again in this one.

The page does not blend these into a 0 to 100 score. A blend needs weights, and no official weights exist. Plain ratios show which part moved.

Rates built from action totals answer a different question. Use the X engagement rate calculator for those.

1. The tool divides repeat engagers by all accounts that engaged and multiplies by 100.
2. It subtracts repeat engagers from all engagers to get one-time engagers.
3. It divides repeat engagers by followers and multiplies by 1,000.
4. It divides all engagers by followers for the engaged share of followers.
5. It divides returning accounts by the earlier period total for the return rate.
6. It subtracts your earlier share from the current share and shows the gap in points.

Repeat engager share looks inside one period. Return rate looks across two periods. An account can count in both.

Example values only. An account with 12,000 followers tallies 20 posts from the last 28 days. 400 different accounts engaged, and 80 of them engaged on 2 or more posts. The earlier 28 days had 350 engagers, and 105 of those appear again. The earlier repeat engager share was 18%.

• Repeat engager share = 80 / 400 x 100 = 20.0%.
• One-time engagers = 400 - 80 = 320, or 80.0% of engagers.
• Repeat engagers = 80 / 12,000 x 1,000 = 6.7 per 1,000 followers.
• Engaged share of followers = 400 / 12,000 x 100 = 3.3%.
• Return rate = 105 / 350 x 100 = 30.0%.
• New engagers = 400 - 105 = 295, or 73.8% of engagers.
• Change on earlier share = 20.0 - 18 = +2.0 points.

One engager in five came back for a second post. Seven in ten of the earlier engagers, 245 accounts, did not appear this period. Both facts are true at once, and a single score would hide one of them.

How Many Posts You Tally

More posts give each account more chances to appear twice. A tally of 40 posts shows a higher repeat share than a tally of 10 posts from the same account. Keep the number of posts equal between periods.

Which Actions You Count

Replies, reposts and quotes are public and easy to list. X Help states that bookmarks are private and only visible to the person who saved the post, so nobody can tally bookmark users. Decide which actions count, write the rule down, and reuse it.

A Post That Travels Far

A widely shared post brings many first-time engagers. They raise the bottom of the ratio, so repeat engager share falls even when no regular left. Read the one-time count next to the share.

Views Are Not People

X Help states that view counts are not unique: one person who opens a post on web and on a phone adds two views. Views therefore cannot stand in for a count of people.

Follower Count Changes

The per 1,000 followers figure moves when followers rise or fall. Track follower change itself with the X follower growth calculator.

Repeat engager share measures loyalty inside one period, and return rate measures it between two periods. An account with a low share can still have a high return rate.

Follower retention is a third measure. It tracks who stays followed, not who engages. The X retention rate calculator covers it.

Monthly Account Review

Tally the same number of posts each month. Record repeat engager share and return rate side by side.

Testing a Series

A recurring format, such as a weekly thread, aims to bring people back. Tally only the posts in the series and compare with your other posts.

Before a Launch

Repeat engagers are the people most likely to see a launch post. The count per 1,000 followers shows how large that core group is today.

Reporting to a Client or Sponsor

Report the ratios with the period, the number of posts and the counting rule. Creators who also run a video channel can compare with the YouTube community loyalty calculator.

1. Trusting a Loyalty Score

A score out of 100 hides its weights. X publishes none, so every such score rests on choices its author made.

2. Counting Actions as People

200 replies can come from 30 accounts. Count handles, not notifications.

3. Comparing Periods of Different Length

A 90 day tally gives every account more time to return than a 28 day tally. Use equal periods and an equal number of posts.

4. Comparing with Other Accounts

X publishes no average repeat share. A figure quoted for another account used another counting rule. Compare with your own earlier period.

5. Counting Your Own Replies

Your own account appears on every thread you answer. Leave it out of both counts.

6. Reading a Falling Share as Lost Fans

The share falls when new people arrive. Check whether the repeat engager count itself fell before you change anything.

Each ratio is exact for the counts you enter. The counts depend on how carefully the tally was made and on which actions you can see.

A hand tally of 10 to 20 posts is a sample of your activity. Treat small moves of 1 or 2 points as noise.

No. X reports counts such as impressions, likes, replies, reposts, quotes and bookmarks. It publishes no loyalty score, so any score shown elsewhere uses weights its author chose.

Count people instead of actions. Tally the different accounts that engaged with a set of posts, then divide the accounts seen on 2 or more posts by the total and multiply by 100.

A repeat engager is an account that engaged with 2 or more of your posts in one period. You can set a stricter rule, such as 3 or more posts, as long as you keep the same rule every time.

X publishes no good or average figure. The share depends on how many posts you tally and how far they travelled. Compare your result with your own earlier period.

Use a fixed number that you can repeat, such as your last 10 or 20 posts. The same number in every period matters more than the size of the number.

No. X Help states that bookmarks are private and only visible to the account that saved the post. You see a bookmark total and no names.

A popular post adds many first-time engagers to the bottom of the ratio. The share falls even when the number of repeat engagers stays the same.

No. Return rate counts earlier engagers who engaged again. Follower retention counts followers who still follow you. A person can keep following and stop engaging.

This page makes no claim about that. X Help says recommendations use a variety of signals and publishes no weights, so the calculator does not guess them.

No. The maths runs in your browser and nothing is sent to us. Save keeps a result only in this browser.

• X API documentation: Metrics (X developer documentation, read 2026-10-03; the post metrics X reports and their definitions, all of which count actions).
• View counts (X Help Center, read 2026-10-03; views are not unique and one person can add more than one view).
• About Bookmarks (X Help Center, read 2026-10-03; bookmarks are private to the account that saved them).
• How to Repost (X Help Center, read 2026-10-03; what a repost and a quote are).
• About your For you timeline on X (X Help Center, read 2026-10-03; recommendations use a variety of signals; no weights are published).