Why most Pump.fun coins never graduate

The graduation rate on Pump.fun is low, and it gets quoted as evidence that something is wrong with the format.

Look at the denominator first. It explains most of the number.

The structural reason

Launching costs almost nothing.

That's the entire design, and it's genuinely useful, because a creator with an idea and 0.02 SOL can test it in 4 minutes rather than raising money and hiring an auditor. The barrier is gone.

Consequence of removing a barrier: everything comes through. Tests, jokes, duplicates, coins minted to see what the button does, and dozens launched per hour by people running the same template. Most mints were never attempts at anything, so measuring them as failures misreads what happened.

Where coins actually stop

Two distinct populations, and they get lumped together.

The first 10%. Overwhelmingly the largest group. These coins have a handful of trades, sometimes only the creator's own, and nothing ever happens. They didn't fail in any interesting sense. Nobody encountered them.

The 70s and 80s. Much smaller, and genuinely interesting. Real buying happened, real holders exist, and then progress stops or reverses. Reading curve progress covers why that band specifically.

Those 2 groups need completely different explanations, and advice aimed at one is useless for the other.

Why the first group stalls

Exposure, almost always.

Browsable surfaces rank by recent activity, so a coin with no trades doesn't rank low, it's absent from the views where strangers discover things. Nobody rejected it, because nobody was shown it.

That's the honest diagnosis for most dead coins, and it's testable. Check unique wallets across the coin's entire life, not just today. Under 40 means exposure. Reading a stalled launch covers the full check.

Why the second group stalls

Different problem, and it's arithmetic rather than obscurity.

By 80%, early buyers are sitting on large multiples and thinking about exits, while new buyers are looking at the most expensive tokens the curve will ever sell. Those 2 populations meeting is exactly what a stall looks like.

Add the sell pressure from anybody who bought in the first minute and the bar can go backwards without a single thing having gone wrong. Can a coin lose graduation progress covers that mechanism.

What the graduating coins have in common

Not tokenomics, since supply and allocation are identical on every Pump.fun launch. Supply and tokenomics covers why that's true by design.

Three things instead.

They were found. Enough people encountered them that a real holder base formed rather than 8 addresses.

They survived the check. Distribution that doesn't scare anybody, a creator wallet without a trail of dead launches, a page that reads as finished. How holders check your dev wallet covers the 60 seconds.

Something about them transmitted. A concept somebody would mention in a group chat, which is how coins actually spread. How to write a coin narrative covers writing for that.

What this means for you

The failure rate isn't a prediction about your coin.

It's a statement about a population dominated by mints that were never trying. What determines your outcome is whether enough people encounter the coin and whether it holds up when they do, and both of those are things you affect rather than things that happen to you.

Start by finding out which group you're in. Unique wallets over the coin's life answers it in 30 seconds, and the answer decides everything you'd do next. The console prices sustained activity against a specific window if exposure turns out to be the gap, with the full cost shown before anything is signed.