Why Most Products Never Deserved a Company

The number everybody quotes is folklore

The “90% of products fail” line has been passed between decks for two decades without a source attached. The peer-reviewed work puts the real failure rate of launched consumer products at roughly 40% or less (Castellion & Markham, 2013). We keep that correction written into our own scoring instructions, because a framework that starts from a myth will score every product as doomed and therefore score nothing at all.

Forty percent is not comforting. It means a well-run launch is a coin flip with a slightly weighted coin — a base-rate bet, not a lottery ticket. That distinction matters: lotteries reward volume, and bets reward selection.

Effort is not the variable

The failed products we see were not built by lazy people. They were built by people who worked hard on something that could not have worked. The costing was wrong before the first ad ran. The category returned a fifth of every order. The supplier could not double output without a capital wall. The benefit could not be photographed, so paid social had nothing to sell.

None of that is an execution problem. All of it is knowable before launch, and almost none of it is knowable by looking at the product and liking it.

Why the gamble persists anyway

Because the alternative is boring. Scoring a product across 82 factors takes longer than a founder’s intuition and produces answers people do not want. The red fields tend to appear in unit economics, which is the part everyone assumes they will fix later with scale. They do not fix it with scale. In direct-to-consumer, growth is paid for out of margin — if the margin is missing, scale buys a larger loss.

That is why our framework treats the four economic factors as close to a knockout rather than as one input among many. Two or more red fields there and the evaluation stops, however green the rest of the sheet is.

What “deserved a company” means

A company is a bet on repetition: that the product sells again, at a price that covers what it costs to find the next buyer, for long enough to be worth building around. Most products cannot carry that. They are fine products. They are not companies.

Our job in Phase One is to find the few that are — and to say no, early and in writing, to the rest. Nobody thanks you for a no. They thank you for the one you gave them in three business days instead of eighteen months.

Read next: What 82 factors actually check. Or put a product through it — apply to Y-Score.