How to Check a List of 50 Crypto Influencers Before You Pay (From an Agency or Your Own)
We ran a 50-name crypto KOL list through four checks: real reach, real replies, audience fit and fair price. 18 could not be checked; 6 of the other 32 passed everything. Here is the method.

CTO & Co-founder at Oznaka

The fastest way to check a list of 50 crypto KOLs is to run every handle through the same four questions — is the audience that sees the posts real, are the replies from real accounts, is that audience your audience, and is the quote fair for the real reach — then sort the list by the answers instead of by follower count. Done by hand it is about five minutes per account, so a 50-name list is a working day. Pasted into Oznaka it is one message and a few minutes, and the output is a sorted table you can send back to the agency.
Oznaka (oznaka.ai) by Wallchain is a tool that finds and vets crypto influencers on X by real reach, real replies, audience fit and fair price.
This guide covers the three situations we hear most often on calls: a list that came from an agency with prices attached, a list you collected yourself from Telegram groups and timelines, and a single handle with a price that you want to sanity-check before replying.
To show what the process finds, we ran a typical agency-style list of 50 crypto KOLs through it on November 16, 2026. 18 names could not be checked at all. Of the 32 that could, 6 passed every check, 18 needed questions before paying and 8 should be dropped. The full breakdown is in the example at the end.
I have a list of 50 crypto KOLs from an agency (or that I collected myself). How do I check which of them are worth paying?
Start by making the list checkable. Agency lists arrive as display names, Telegram usernames and screenshots; your own list is usually a spreadsheet with half the handles typed from memory. Before any vetting, turn every row into an X handle, remove duplicates (the same creator often appears twice under an old and a new name), and keep a column for where the name came from and the quoted price, if there is one. You will need the price later to compare it with real reach.
Then ask the four questions, in this order, because each one removes names and the later checks are slower.
1. Does the audience that sees the posts exist? Follower count tells you nothing on its own: followers can be bought, inherited from a sold account, or dead since 2022. What matters is how many of them actually see a post. Compare median views on the last 20–30 posts with the follower count. A real account with 60K followers usually shows views in the low thousands to low tens of thousands per post; a washed account shows either suspiciously round, flat view counts on every post or views that are a tiny fraction of followers. Sort out everyone whose median views do not match their size. In Oznaka this is two columns: Impr. Rate, the average share of followers who see a post, and Org. Impr., the estimated share of those impressions that are organic rather than boosted.
2. Are the replies from real accounts? Open five posts and read the replies, not just count them. Reply farming looks the same everywhere: the same thirty accounts under every post, one-line comments that could be under any post ("great thread ser", "bullish"), accounts created in bulk in the same month, and a conversation that never continues past the first reply. Real audiences argue, ask questions, quote-tweet and disagree. If the replies read like a Telegram task list, the engagement is being manufactured and the views are probably assisted too. In Oznaka this is Org. Replies.
3. Is the audience your audience? This is the check most agency lists fail, even when the accounts are real. A creator with a real 40K audience of NFT traders is a bad buy for a perp DEX, however clean the numbers are. Look at what the audience actually engages with: which projects they reply to, which topics get traction on this account, what the last ten quote-tweets were about. A list of 50 "crypto KOLs" usually contains three or four niches mixed together, and only one of them is yours. Oznaka's Crypto Focus and Crypto Share show how much of an account's recent posting is about crypto at all; if most of a list fails here, it is faster to build your own shortlist from your audience.
4. Is the quote fair for the real reach? Once you know real views and the share of the audience that fits, divide the price by the number of real, relevant readers rather than by followers. A $1,500 post that reaches 20,000 real readers in your niche is cheaper than a $400 post that reaches 1,500. Agency lists are priced by follower tier, which is exactly why they look expensive once you redo the math on reach. Oznaka shows an estimated price and CPM for every account, and our survey of what 1,051 crypto creators on X charge has benchmarks per tier.
After the four questions, sort the list into three buckets: keep (passes all four), ask (passes reach and replies but the fit or price is unclear), drop (fails reach or replies — no price fixes a fake audience). On the list we ran for this guide, that was 6 keep, 18 ask and 8 drop out of 32 checkable accounts, plus 18 names that could not be checked at all.

Loading embed…
How do I tell if the replies under a crypto influencer's posts come from bots or reply-farming Telegram groups?
Reply farming is the hardest signal to fake well and the easiest to spot once you know the pattern, because the farm has to be cheap to run. Check five things on five recent posts.
Who replies. Open the profiles of the first ten replies. Farmed accounts tend to have no original posts, a profile created within a few weeks of each other, a follower count under a hundred, and a timeline that consists only of replies to a dozen creators. Real repliers have their own posts and their own followers.
What they say. Farmed replies are generic by design because one person writes thirty of them. Any reply that could be moved under a different post without edits counts as generic. A real audience responds to the content of the post.
Whether the same accounts recur. Scroll back five or ten posts. If the same names appear in the first replies every time, you are looking at a roster, not an audience. Most reply pods and Telegram engagement groups have a fixed membership, and the repetition is visible without any tool.
Timing and depth. A burst of replies in the first two minutes followed by silence is the signature of a group that was pinged; real engagement trails over hours. And real replies get replies: a post with 80 top-level comments and zero second-level ones was commented on by people who were not reading it.
Oznaka reports this as Org. Replies — the estimated share of replies that are genuine rather than paid or farmed — so you do not have to open profiles by hand on a 50-name list. On our test list, 27 of the 32 accounts had a replies reading. For 9 of them, fewer than 70% of replies looked genuine, and for 3, fewer than 40%.
Loading embed…
A crypto influencer with 63K followers and an influence score of 240 sent me a price for a post. Here's the handle — is the audience real?
The honest answer is that neither number tells you. 63K followers is a size, not a quality; an influence score from a ranking site is mostly a function of who follows the account, which is useful for seeing whether other creators know them and useless for seeing whether their followers read them. The two checks that answer the actual question are the ones above: median views against followers, and who is in the replies.
For a single handle, do this: open the last 30 posts and write down the views on each (skip pinned and viral outliers), take the median, and compare it with 63K. If the median is in the 3–15K range and the replies are from real people, the audience is probably real and the question becomes fit and price. If the median is in the hundreds or the views are identical on every post, the score was never telling you anything. Then look at who the repliers are and what they engage with elsewhere — this is the fit check, and it is where a handle with a perfectly real audience can still be the wrong buy for your product.
If you want this done in one step, paste the handle into Oznaka and read the columns together, not one at a time. On our test list, the account with the highest impression rate — about 48% of followers seeing an average post, nine times the list median — was also the one where Oznaka estimated under 1% of those impressions as organic and about a quarter of the replies as genuine. At the other end, seven accounts with an X Score above 300, meaning many credible accounts follow them, reached fewer than 4% of their followers on an average post. Neither number alone told the story; reach, organic impressions and organic replies together did.
Loading embed…
What a bad agency list looks like
Agencies are not the villain here: many run honest lists. But a few patterns should make you check harder rather than less.
The same roster for every client. If the list you got looks like the list a friend got for a different product, the agency is selling the creators it has relationships with, not the ones your audience follows. Fit is the first thing to check.
Prices in neat tiers. $500 / $1,000 / $2,500 by follower band means the list was priced from a spreadsheet, not from reach. Redo the price per real reader and expect the ranking to change.
Packages. "10 KOLs for $15K" bundles strong accounts with filler. Ask for the per-creator price and vet each one; the package usually exists to move the filler.
No view numbers. An agency that quotes followers but not views either has not looked or does not want you to. Ask for median views on the last 20 posts; a good agency has them.
Agency margins are normal. An agency adds its margin on top of what the creator receives; that is its business model, not fraud. Ask for the per-creator price so you can compare like with like (more on how KOL pricing works). A margin only becomes a problem when it sits on top of a fake audience, which is why the reach check comes before the price check.
What to send back to the agency
A sorted list is also a negotiating position. Send the three buckets with one line of reasoning per dropped name ("median views 400 against 58K followers", "same 20 accounts in every reply thread"). Ask for replacements for the drop bucket in the same niche as your keep bucket, ask for per-creator prices on the ask bucket, and send your brief again so the replacements fit. On our test list, only 6 of the 50 names would have gone straight to a brief; the other 44 needed a replacement, a question or an explanation.
Skipping this step is how a campaign ends up with millions of views and almost no sign-ups.
Example: a 50-name list, before and after
On November 16, 2026 we built a list the way an agency would: 50 handles taken from public "top crypto influencers on X" pages published by crypto marketing agencies and media, with big names, mid-sized accounts and small ones mixed together. We pasted it into Oznaka with one line of context, a perp DEX launch. Oznaka matched 32 of the 50.
How we sorted. A check passed when Oznaka showed it in its green or yellow band and failed in orange or red; a missing reading counts as not passed. Reach is Impr. Rate (the share of followers who see an average post); organic impressions and organic replies are Org. Impr. and Org. Replies. Oznaka has no perp-specific fit signal for a pasted list, so for fit we used the closest proxy: crypto is at least about a third of recent posts (Crypto Share). Keep: all four passed. Drop: reach, organic impressions or organic replies in the red band. Ask: everything else. This is one list on one day, so read the numbers as an example of the process, not a market-wide rate. No handles are published.
| Followers | Checked | Reach passed | Organic impressions passed | Organic replies passed | Crypto fit | Keep / ask / drop | Median est. price per post | Est. price per 1,000 organic views (median) |
|---|---|---|---|---|---|---|---|---|
| Under 50K | 11 | 8 | 6 | 3 | 9 | 1 / 3 / 7 | ~$90 | ~$102 |
| 50K–250K | 7 | 6 | 5 | 4 | 6 | 4 / 3 / 0 | ~$453 | ~$78 |
| 250K+ | 14 | 5 | 11 | 11 | 11 | 1 / 12 / 1 | ~$1,087 | ~$50 |
| All checked | 32 | 19 | 22 | 18 | 26 | 6 / 18 / 8 | ~$508 | ~$78 |
| Not in Oznaka's index | 18 | — | — | — | — | — | — | — |
18 of the 50 names could not be checked. Oznaka's index covers individual creators with at least 2,000 followers, an X Score of 50 and recent original posts, so this group mixes accounts that have not posted original content recently, accounts with too few credible followers, brand accounts and creators the index has not covered yet. We left all 18 out of the numbers rather than guess. On a real list, they are the first thing to ask the agency about.
Small accounts failed most often. 7 of the 11 accounts under 50K followers landed in drop: five on organic impressions or organic replies, two because their posts reached almost none of their followers. Two of the three accounts with the highest impression rates in this group were among the drops: for one, Oznaka estimated almost none of the impressions as organic; for the other, most of the replies as not genuine.
Mid-sized accounts were the strongest part of the list. 4 of the 7 accounts with 50K–250K followers passed every check, and none was dropped.
Big accounts mostly landed in ask, not drop. 12 of the 14 accounts with 250K+ followers needed questions. Their impressions and replies looked mostly organic (medians of 73% and 75%), but the median account reached about 4% of its followers on an average post. A typical big account here had about 629,000 followers and about 25,000 views per post. That does not have to mean anything is fake, since large, older audiences are often mostly inactive. It does mean the follower count overstates the audience about 25 times, and a quote priced by follower tier needs to be re-priced from views.
Price per follower and price per real view tell different stories. Priced per follower, the median account on the list costs about $2.80 per 1,000 followers. Priced per organic view, about $78 per 1,000. By that measure the big accounts were the cheapest (about $50 per 1,000 organic views) and the small ones the most expensive (about $102), because so much of the small accounts' reach was not organic. These are Oznaka's estimated prices per post, not the agency's quotes; with real quotes, run the same division.

FAQ
Check your list now. Paste your list into Oznaka and get real reach, real replies, audience fit and an estimated price for every account it can match.
Find creators who fit your next campaign.
Start with your audience, your niche, and the outcome you want to achieve.
Find creators

