The volume bot mistakes that waste the most sessions are sizing errors: a target too thin for the board at that hour, one short burst followed by silence, and an order band so narrow that every trade looks alike. Next come setup slips, such as a wrong mint or an unfinished token page, and expectation errors, such as counting on a price move or never checking results on-chain.
If you only fix three
- Size the session for the hour you run it, and spread it so the activity is continuous rather than a spike.
- Give the order band room: a max two to four times the min gives a visible mix of order sizes.
- Check the mint twice and finish the token page before paying, because the fee is not refundable once a session starts.
- Note numbers before and after, then judge the session by what settled, not by the price.
Is your session too thin for the board?
Undersizing rarely means the total is low. It usually means the activity per minute is thin for how busy the board is at that moment. Pump.fun board activity rises and falls with everyone else launching at the same time, and a session competes with all of it.
Take the 100 SOL minimum: 500 wallets at an average of 0.20 SOL, about 800 orders. Over 60 minutes that is roughly 13 orders a minute. Stretch the same session over 10 hours and it drops to about 1.3 a minute, a pace that tends to vanish among other coins. Fix: match length to size. Keep small sessions short, or add wallets or lift the band before stretching the clock. The signals visible on the board are worth reading at the hour you plan to start.
Why does one big burst fall flat?
The opposite error packs a large session into a few minutes. Picture the Big prints shape, 1,250 wallets at 0.40 to 1.20 SOL, run over 15 minutes instead of 2 hours. The average order is 0.80 SOL, the target is 1,000 SOL and the fee is 10 SOL. About 2,000 orders land in a quarter hour, and then the tape goes quiet.
A chart with one spike and a flat line after it tends to raise questions rather than attract buyers, and visitors who arrive twenty minutes later see nothing happening. Fix: keep the preset length, or split the budget into two sessions with a gap between them, so activity covers the window when your community is actually watching.
What goes wrong with identical order sizes?
Every order size is drawn at random inside the band you set, so the band decides how varied the tape looks. A band of 0.24 to 0.26 SOL and a band of 0.10 to 0.40 SOL both average 0.25 SOL and give the same target, but the second spreads orders fifteen times wider. The narrow one produces trades that all read as the same size to anyone scrolling the trade list.
Fix: set the max at two to four times the min. The presets sit in that range: First hour and Quiet tape at 3x, Slow build at 4x, Big prints at 3x. There is more on picking the numbers in the order size guide.
Which setup slips happen before the first order?
These three cost nothing to prevent and cannot be repaired once the session is running.
Pasting the wrong mint
A mint copied from a chat, a screenshot or a lookalike coin sends the whole session to the wrong token. PumpCrest reads the name, ticker and image for the mint you paste so you can confirm it. Fix: copy the address from the live coin page, compare the first and last four characters, and look at the image the console shows before paying. Orders that settle on the wrong token stay on-chain.
Running before the token page is ready
The session brings attention to the coin page. If the image is a placeholder, the description is empty or the Telegram link is dead, that attention leaves. Fix: finish the page and socials first; the pre-launch checklist gives each item a deadline.
Comments that do not fit the token
Library lines in a language your community does not speak, or generic hype under a coin with a specific theme, read as noise. Fix: choose the library language your holders use, or upload a .txt file of your own lines, up to 512 KB and read only in your browser. Keep density sensible, since 20% of about 1,600 orders is already around 320 comments.
Which expectations set a session up to disappoint?
Expecting a price outcome
Presets lean toward buys, between 58% and 68% of orders, but sells still run and every other trader on the coin moves price too. PumpCrest reports volume, transactions, holders, comments and favorites as they settle; it does not promise a price move, a trending rank or a profit. Fix: set goals you can measure on-chain, such as holder gain or makers inside the session window.
Ignoring the hour of day
A session that runs while your community sleeps produces activity few of your own people see, and board conditions shift through the day. Fix: work out where your holders are, start when they are online, and look at the board at that hour before you pay. Reach for the long quick picks, 6h or 10h, mainly when your community spans several time zones.
What do you lose by skipping the on-chain check?
Without before-and-after numbers you cannot tell whether a setting worked, so the next session repeats the same guesses. The check takes about ten minutes: the Solscan link for your fee payment, the token's holder count and transfers, and volume, transactions and makers on DexScreener once a pool exists. Fix: note a baseline before you start and verify the results on Solscan the same day. Compare the holder gain with the console estimate of about 62% of wallets, keeping in mind that it is an estimate.
All nine mistakes side by side
| Mistake | What happens | Fix |
|---|---|---|
| Undersizing for the hour | Activity per minute is too thin to stand out | Shorter length or more wallets; read the board first |
| One burst, then nothing | A spike followed by a flat tape | Keep the preset length or split into two sessions |
| Narrow order band | Every trade reads as the same size | Max at 2x to 4x the min |
| Wrong mint pasted | The session runs on the wrong token | Copy from the coin page; check image and characters twice |
| Page not ready | Visitors arrive and leave | Finish image, description, links and socials first |
| Comments that do not fit | The thread reads as noise | Right language, or upload your own .txt lines |
| Expecting a price outcome | The session is judged by the wrong number | Measure holders, makers and transactions instead |
| Ignoring the hour of day | Activity your own community never sees | Start when your holders are online |
| No on-chain check afterward | No baseline for the next run | Log numbers before and after; verify the same day |
A one-minute pre-flight before paying covers most of the table:
- The mint matches the coin page and the image is confirmed.
- The band max sits between 2x and 4x the min.
- Orders per minute make sense for the length you picked.
- The comment language matches your community.
- Baseline numbers are written down with a time.
Copy the shareable config link from the console and send it to a teammate before paying. A second pair of eyes catches a wrong mint or an odd band in seconds.
Questions and answers
Which mistake is the most expensive?
The wrong mint. Nothing in a session can be redirected once orders settle, and the fee is not refundable after the start. Sizing errors waste part of the value; a wrong mint wastes all of it. The console shows the name, ticker and image for the address you paste, so a careful look at that preview is the cheapest protection there is.
Can I fix a mistake while the session is running?
Partly. To stop or change a running session you message support, and only what has not run yet can change, because settled orders stay on-chain. That helps with a session that is too long or a comment density you want lower. It does not help with a wrong mint or a page that should have been ready earlier.
Is a bigger budget the fix for most of these?
Rarely. At the flat 1% rate, doubling the target from 250 to 500 SOL doubles the fee from 2.5 to 5 SOL, but a bigger session run at the wrong hour, with a narrow band or on an unfinished page, carries the same problems at twice the price. Fix the shape first, then scale it.
How do I know if my band is too narrow?
Divide the max by the min. A ratio near 1, such as 0.20 to 0.22, means nearly identical orders, while a ratio between 2 and 4 matches the built-in presets. You can also scroll the trade list on the coin page in the first minutes: if every line shows nearly the same amount, widen the band next time.
What is a sensible first session that avoids most of these?
The First hour preset: 500 wallets, a 0.10 to 0.30 SOL band and 60 minutes. It averages 0.20 SOL per order, lands on the 100 SOL minimum for a 1 SOL fee, runs about 800 orders at roughly 13 a minute and uses a 3x band. Run it while your community is online, after the page is finished, and check the results the same day.
Set up a session without the usual slips
The console previews your mint and shows band, length and fee before you pay, and a copied config link lets a teammate review the setup first.
We build the PumpCrest console and write these guides from the operator side: what a session spends, what settles on-chain and what a visitor sees on the board. Corrections are welcome at support@pumpcrest.com.