How to spot fake Twitter followers on a crypto project comes down to comparing 4 numbers rather than reading 1: the follower count against an engagement or influence score, the account's age, the bio against the bios of the accounts around it, and the follower count against its own earlier value. NAUT tracks 10,583 live project rows, 8,812 of them with a stored X follower count and 2,275 with an X intel snapshot that carries an influence and heat score next to that count, so on those 2,275 rows 3 of the 4 checks can be run from stored data before anyone opens a browser.
This article runs the 4 checks on 1 of those rows, Pixelverse, a Telegram tap-to-earn game with the $PIXFI token, and shows what each check returns. Pixelverse is a funded project, not an accusation: NAUT's own record puts Bitscale Capital among the backers of its last tracked fundraise, dated 12 June 2024. The point of the 4 checks is to tell you which question to ask next, and on this row they produce 1 clear question.
Data note. Every count here was read from NAUT's own database on 8 November 2026, read-only. The Pixelverse figures are dated: the stored follower count of 2,484,770 was last checked on 30 May 2026, the X profile read behind it was fetched on 6 June 2026, and the X intel snapshot was imported on 1 April 2026. Treat all 3 as last tracked by NAUT, never as a live count. 5 claims about the public X profile were put to outside sources: 3 came back confirmed and are used below, and 2 came back contradicted and are left out of this article. NAUT does not store X bio text for company rows, so check 3 carries a method rather than a number. NAUT Research wrote it.
How do you spot fake Twitter followers on a crypto project?
Run the 4 checks in the order below, because each one removes an innocent explanation for the one before it. Together they are the X follower count check that crypto project due diligence usually skips. A large follower count on a young account is one story, the same count on a 3-year-old account is another, and the same count that appeared in under 2 months is a third.
| Check | As last tracked | What it flags |
|---|---|---|
| 1. Ratio | 2,484,770 followers, influence 118, heat 14 | A low score for that size |
| 2. Age | Joined 26 October 2023, 1,109 days old | Age does not explain the count |
| 3. Bios | Not stored, read on x.com | Method only, no figure |
| 4. Spikes | 39,318 on 1 April 2026, 2,484,770 on 30 May 2026 | A 63.2x jump in 59 days |
Ranked by how much each check costs you, the order is cheap to expensive: checks 1, 2 and 4 are 1 NAUT page or 1 API call on any of the 2,275 rows with an intel snapshot, and check 3 is manual reading on x.com for every row.
Check 1: does the follower count match the engagement score?
A bought follower does not click. That is the whole basis of this check, and it is why an influence or engagement score sits closer to the truth than a follower count does.
NAUT's X intel snapshot for Pixelverse carries an influence score of 118 and a heat score of 14 beside a follower figure of 39,318, all imported on 1 April 2026. The stored follower column for the same row reads 2,484,770, last checked on 30 May 2026. Scores of 118 and 14 are low for an account carrying millions of followers, and that is the mismatch this check exists to surface.
2 supporting ratios come from the same row. Pixelverse follows 136 accounts against those 2,484,770 followers, a ratio of 18,270 followers for every account it follows, and it has posted 686 times, which is 3,622 followers for every post. Neither figure is proof of anything on its own. A real game with a viral Telegram mini-app will show both patterns. They tell you the account's reach was not built post by post.
Check 2: how old is the account?
A 4-week-old account with 2,000,000 followers answers itself. An old account does not, which is why age is a filter rather than a verdict.
Pixelverse joined X on 26 October 2023, which is 1,109 days, or 3 years, before 8 November 2026. An outside read of the public profile confirmed that join date, and the account's numeric X id decodes to the same timestamp. So age is not the explanation here. The account had 2 and a half years of runway to build an audience, and the question moves to when the audience actually arrived, which is check 4.
Check 3: is the bio duplicated across accounts?
This is the 1 check with no stored figure behind it. NAUT does not keep X bio text on company rows, so there is no Pixelverse number to print and no count to quote. It is a manual read, and it is still worth the 10 minutes.
Open the account, then open 10 to 20 of its recent followers. What you are looking for is repetition: the same sentence in 8 bios, the same emoji string, the same link, default avatars, handles that run name plus 8 digits, and join dates that cluster in 1 month. Follower farms are built from templates, and templates repeat. Do the same on the project's own bio: a bio that copies another project's wording word for word is a different warning, and it costs 1 search to find.
Record what you see as a date and a count in your own notes, because the accounts you looked at will be gone or renamed in a quarter.
Check 4: did the follower count spike?
2 dated readings of the same number are worth more than either reading alone, and this is where the Pixelverse row produces its question.
NAUT holds 39,318 followers in the snapshot imported on 1 April 2026 and 2,484,770 in the follower column checked on 30 May 2026. That is 63.2 times the earlier figure in 59 days. 2 readings explain it, and NAUT's data does not settle which one is right. Either the account really did gain that audience in under 2 months, which a Telegram mini-app push, an exchange listing or an airdrop campaign can do to a tap-to-earn game, or the 39,318 figure was already stale at the moment it was imported and the snapshot was describing an older state of the account. The honest version of this check is that the gap is real and its cause is open.
What a spike never means on its own is fraud. What it does mean is that any number you quote from that account needs its own date attached, and that an influence score imported alongside the smaller figure cannot be read against the larger one.
- Pull the 2 follower readingsPixelverse holds 39,318 from 1 April 2026 and 2,484,770 from 30 May 2026, a 63.2x gap.
- Read the scores beside themAn influence score of 118 and a heat score of 14 are low for millions of followers.
- Date the accountPixelverse joined X on 26 October 2023, so at 1,109 days old its age explains nothing.
- Read 20 follower biosNAUT stores no bio text for company rows, so this one is a manual read on x.com.
- Attach a date to every numberOf NAUT's 10,583 live project rows, 8,812 carry a stored follower count with a check date.
What the 4 fake Twitter followers checks tell a BD team
A follower count is a sourcing input, and these 4 checks decide how much weight it carries. Run them as your fake Twitter followers crypto project screen before an account's reach goes into a target list.
- Rank by the score, not the count. On the 2,275 NAUT project rows with an X intel snapshot, the influence and heat scores sit next to the follower figure, so a sort by reach can be sanity-checked in the same view.
- Treat 2 dated readings as the signal. A 63.2x gap between 1 April and 30 May 2026, as on the Pixelverse row, is the kind of move worth a question in the first call.
- Price the account age in. An account 1,109 days old with a flat history and a sudden jump is a different conversation from a 3-month-old account with the same count.
- Do not pitch off a stale number. Every Pixelverse figure above was last tracked between 1 April and 6 June 2026, which is why each one here is printed with its date.
- The 1,771 live project rows with no stored follower count are not failures of the project. They are rows where this check has to start on x.com instead.
What a founder raising should take from this
If you are raising, an investor will run some version of these 4 checks on your account before the second call, and the cheapest 3 of them run off stored data.
- Know your own 2 readings. If your follower count moved by a multiple in weeks, have the campaign, listing or mini-app push that caused it ready as a sentence.
- Bring engagement, not reach. An influence score of 118 against millions of followers invites the question. Impressions, retention and wallet counts answer it.
- Keep 1 handle. NAUT matched the Pixelverse row to its CoinGecko listing by its X handle, which is how a single handle across your site, docs and token listing keeps your own record clean.
- Check what a tracker holds on you. NAUT's stored follower figure for Pixelverse carries a check date of 30 May 2026, and a figure that old is the one an investor will read first.
Related reading on NAUT: Crypto contact data accuracy: 86 of 100 X handles, 0 suspended and How to verify a crypto investor: 6 checks, 143 funds pass 4.
Can you pull this data through the NAUT API?
Yes, for the 3 checks that run off stored data. GET https://api.naut.so/v1/projects/pixelverse-2 returns the Pixelverse company record with its team, which is where the stored X handle and follower figures behind this article live, and GET https://api.naut.so/v1/projects?q=tap%20to%20earn&token=launched lists comparable rows to screen the same way. Every call is a GET with an Authorization: Bearer $NAUT_KEY header and costs 5 credits, documented at naut.so/docs/api with plans from $49 a month for 3,000 credits at naut.so/pricing. Verified Telegram, email and LinkedIn for the people on these rows unlock on any paid NAUT plan, so start a 7-day trial if you need the channel and not only the number.
Fake Twitter followers FAQ
How do you check for fake Twitter followers on a crypto project?
Compare the follower count against an engagement or influence score, check the account's join date, read 10 to 20 follower bios for repeated text, and compare the count against an earlier dated reading of itself. On Pixelverse on 8 November 2026 those 4 checks returned 2,484,770 followers against an influence score of 118, a join date of 26 October 2023, no stored bio text, and a 63.2x rise from 39,318 followers between 1 April and 30 May 2026.
Does a sudden follower spike mean a crypto project bought followers?
No. A 63.2x rise like the one on NAUT's Pixelverse readings between 1 April and 30 May 2026 can come from a Telegram mini-app push, an exchange listing or an airdrop campaign, and it can also mean the earlier figure was stale when it was recorded. It is a question to ask on the call, not a finding.
How many crypto projects can these checks be run on from stored data?
NAUT tracks 10,583 live project rows as of 8 November 2026, 8,812 with a stored X follower count and 2,275 that also carry an X intel snapshot with an influence and heat score. Checks 1, 2 and 4 run from stored data on those 2,275 rows, and check 3 is a manual read on x.com for any row.
Where can I find a crypto project's X handle, follower count and backers in one place?
On its NAUT project page and through the API: the Pixelverse page carries the stored X handle, the 2,484,770 follower figure with its 30 May 2026 check date and Bitscale Capital among the backers of its last tracked round, dated 12 June 2024, and GET https://api.naut.so/v1/projects/pixelverse-2 returns the same record for 5 credits. Verified Telegram, email and LinkedIn for the named people on a row unlock on any paid NAUT plan, so start a 7-day trial to reach them.
What does a low influence score next to a high follower count mean?
It means the account's measured engagement does not scale with its audience, which is the single most useful fake-follower signal because bought followers do not click. On the Pixelverse row NAUT's snapshot of 1 April 2026 holds an influence score of 118 and a heat score of 14, both low for the 2,484,770 followers the row stored on 30 May 2026.
Sources
- NAUT's own database, read on 8 November 2026: 10,583 live project rows, 8,812 with a stored X follower count, 2,275 with an X intel snapshot. The Pixelverse row holds 2,484,770 followers checked 30 May 2026, an X profile read fetched 6 June 2026 (686 posts, 136 following), and an intel snapshot imported 1 April 2026 (39,318 followers, influence 118, heat 14). The same row's last tracked fundraise is dated 12 June 2024, with Bitscale Capital among its backers.
- The public X profile at x.com/pixelverse_xyz, with the handle also listed on the project's own documentation at docs.pixelverse.xyz. 3 of the 5 claims put to these sources came back confirmed: the join date of 26 October 2023, a post count of 684 against NAUT's stored 686, and 137 accounts followed against NAUT's stored 136. 2 claims came back contradicted and are left out of this article.