How People Actually Clean Their X Accounts
Usage data from the launch cohort of X Cleaner, the Chrome extension for bulk-deleting tweets, likes, retweets and replies: the first 30 days after public launch, March 25 – April 24, 2026. Every number on this page comes from the extension's own usage counters and our server records — no surveys, no extrapolation.
The three numbers that matter
If you're looking for a single data point to cite, start here.
The first two numbers surprised us. The working assumption for anyone building a "tweet deleter" is that users want to erase their tweets — the public-facing record. In practice, nearly every user in the launch cohort came to scrub their likes: the list of everything they have liked since joining, which stays attached to the account even though X hid likes from other users in 2024.
What gets deleted
Breakdown of content type across all tracked deletions, measured on the first 17 days of the launch cohort (March 25 – April 10, 2026; n = 9,286 items).
Deletion speed in practice
Your X password and session never leave your browser: X Cleaner acts through your already logged-in x.com session, without OAuth or API access. The content it works on (tweets, likes, replies…) is read locally to be displayed and sorted, and is never sent to our servers — see the privacy policy.
The extension works at about one item per second, with built-in delays and automatic pauses when X rate-limits, to limit the risk of being throttled by X. In practice, expect hundreds of items in a few minutes and thousands in about an hour, depending on X's limits. You can pause, resume or stop at any time with a live ETA, and an interrupted clean picks up where it stopped. Each scan loads about 1,000 items per pass (with a "Load more" button), and X only exposes your ~3,200 most recent tweets; older tweets are reached with the Pro archive import, which reads your X archive locally in the browser — nothing is uploaded.
Who uses a tweet deleter — and why
We don't collect demographic data. But we do track behavior patterns from 58 users tracked in the first launch cohort.
| Behavior | Share of users | What it tells us |
|---|---|---|
| Hit the 10/day free cap on first session | 43% | Strong intent: people come to delete a lot, not a little. |
| Never completed a deletion | 53% | Activation friction — typically because the user installed, browsed, and came back later. |
| Returned for a second session within 48h | 43% of first-day cohort | The product has a "finish the job" pull — people come back to complete the clean. |
| Used a date filter | 71% of paying users | "Delete tweets older than X" is the dominant Pro use case. |
| Used keyword filter | 38% of paying users | People target specific past content — usually embarrassing takes, ex-partners, deleted brands. |
Where tweet cleaners come from
Geographic distribution of signups during the first 17 days of the launch cohort (March 25 – April 10, 2026), before geographic ad targeting was tightened.
The geographic distribution contradicts the usual assumption that Western users dominate. The large shares from Indonesia, Brazil and Saudi Arabia suggest the "clean your X account before a job / a marriage / a political shift" use case has strong international demand — often stronger than in markets where the concept was born.
Methodology
Every number on this page comes from one of three sources:
- Extension usage counters. To run the license, the extension reports deletion counts by content type to our server, tied to the X user ID. No tweet content is collected; country is derived from the IP address, which is not stored. This page only shows aggregates — no individual account is identified. Full details in the privacy policy.
- Server records. Signups, license purchases and conversion events.
- Our own Google Ads reports. Geographic distribution of paid traffic, for cross-reference with organic geography.
Period: the launch cohort, i.e. the first 30 days of public launch (March 25 – April 24, 2026). The content-type breakdown, the power-user split and the geographic distribution were measured on the first 17 days of that cohort (March 25 – April 10, 2026). Early-cohort data tends to over-index on power users and tech-savvy audiences. This page is a one-time snapshot of that cohort: it is not updated, and later periods may look different.
Limitations: we don't have perfect visibility into failed deletions caused by X's rate limiting (the extension retries silently and only reports completed operations). Actual attempt counts are therefore slightly higher than reported deletion counts.
License: all data on this page is released under CC BY 4.0. You're free to cite, quote or reproduce it with attribution.
How to cite this data
If you're writing an article, a thread or building on top of these numbers, here are the canonical forms:
X Cleaner. (2026). Usage statistics. Retrieved from https://x-cleaner.app/data
"97.5% of deletions performed via X Cleaner during its launch cohort (March–April 2026) targeted likes, not tweets or retweets (X Cleaner usage data, 2026)."
Source: X Cleaner Usage Statistics — launch cohort (March 25 – April 24, 2026). URL: https://x-cleaner.app/data. License: CC BY 4.0. Methodology: extension usage counters; content-type sample n = 58 users / 9,286 deletions over the first 17 days (March 25 – April 10, 2026). One-time snapshot, not updated.
See your own numbers
Install X Cleaner and watch your deletion count tick up in real time. Free plan includes 10 deletions per day. No login required.
Add to Chrome — FreeRelated
- → Delete all your tweets in minutes — the full pillar guide
- → Delete all your Twitter likes (the top use case)
- → X Cleaner vs TweetDelete — honest comparison
- → Blog: tutorials, deep-dives, changelog