How Pump.fun Holders Grow During a Volume Session

Pump.fun holders are wallets that still have a balance of the token. During a session the count climbs as session wallets buy, and slips whenever a wallet sells its whole position. Because not every wallet ends the run holding, the PumpCrest console shows an estimated unique-holder figure of about 62% of wallets, roughly 620 for a 1,000-wallet session. Treat it as a planning number, then confirm the real count on-chain.

What counts as a holder on a Pump.fun token?

A holder is any wallet with a non-zero balance of the token at the moment you look. The count is a snapshot, not a running total of everyone who ever bought. A wallet that bought at minute five and sold everything at minute forty was a holder for 35 minutes and is not one now.

That snapshot nature explains most of the confusion around holder numbers. Buys add to the count, full exits subtract from it, and partial sells change nothing because the wallet still holds something. On a token that is still on its bonding curve, you will usually see the curve account itself near the top of the distribution, since it holds the unsold supply.

Why does the console show an estimate?

Before a session runs, nobody can know the exact count it will end with. Order sizes are drawn at random inside your band, spacing between orders is uneven, and some wallets will sell their full position along the way. What the console can do is apply one consistent ratio: it shows estimated unique holders at about 62% of wallets in rotation, and labels the figure as an estimate.

Reading it that way matters when you plan. If your team wants the count to clear 1,000, a 1,600-wallet session points to about 992, which falls just short. Dividing the goal by 0.62 gives the wallets you need, 1,613 in this case, and rounding up to 1,700 leaves headroom with an estimate near 1,054.

The 62% figure is a planning ratio applied to every session. It is not a promise; the on-chain count after the run is the number that counts.

Why are holders lower than session wallets?

Two mechanics separate the wallet count from the holder count.

• Sells. Every built-in preset buys more often than it sells, with a buy share between 58% and 68% of orders. Even so, some wallets end on a sell that empties their position, and each of those drops off the holder list.
• Rotation. Wallets take turns placing orders, about 1.6 each on average, so 1,000 wallets produce roughly 1,600 orders but can never add more than 1,000 holders. Whether a wallet counts depends on where its last order left its balance.

Fresh session wallets are created for every run and each is funded with a different amount, so every holder that settles is a separate wallet on-chain rather than one address counted twice.

What do visitors read from the holder list?

The headline count is what people see first, but anyone who scrolls further looks at shape. Most token pages, Pump.fun included, show the largest holders and their share of supply, and a short look at that list tells a visitor a lot.

• Top-holder share. One or two wallets with a large slice of supply reads as concentration, whatever the headline count says.
• The creator's position. Visitors often look for the creator wallet and how much of the supply it still holds.
• Growth over time. Anyone watching for a while notices whether the count rises steadily or jumps in a single burst.
• Holders against trades. Comparing the holder number with the transaction count gives a rough feel for how many separate wallets are involved.

How should holder growth be paced across a session?

Holder growth follows session length. The same wallet count spread over a longer window adds holders at a gentler rate, while a short window packs them together. Dividing estimated holders by session hours gives a simple average pace, which makes setups easy to compare. Choosing the window itself is covered in our guide to how long a session should run.

Because spacing is uneven by design, the real count climbs in irregular steps around those averages rather than ticking up at a fixed rate. To plan a custom session around a holder goal, work backwards:

1. Set a holder goal

   Pick the number you want visible by the end of the run, for example 1,000.

2. Convert it to wallets

   Divide by 0.62. A 1,000-holder goal needs about 1,613 wallets; 1,700 gives an estimate near 1,054 with some headroom.

3. Choose the window

   Decide how long the growth should take. 1,700 wallets over 4 hours averages about 264 new holders an hour; over 10 hours, about 105.

4. Fit the order band

   Set min and max so the target fits the budget. At 0.10 to 0.30 SOL, 1,700 x 0.20 = 340 SOL target and a 3.4 SOL fee at the flat 1% rate.

Does the buy share change how many wallets end up holding?

In practice it can, at the margins. A higher buy share means fewer sell orders, so fewer wallets are likely to exit their full position. The console estimate stays at a flat 62% of wallets, though, so treat any shift from buy share as something you observe after the run, not a figure the console predicts. The wider trade-offs of moving that split are covered in choosing a buy/sell ratio.

How do you check the real holder count afterwards?

PumpCrest reports what settles on-chain, holders included. For an independent read, open the token's mint on Solscan and look at its holders tab, or read the holder figure on the Pump.fun token page. Small gaps between sources are normal, since explorers index at different speeds. Our walkthrough on verifying a session on Solscan goes through the other numbers worth checking, and the Pump.fun volume bot overview explains the full product.

Questions and answers

Written by the PumpCrest team

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.