Twitter went mainstream before anyone worked out the business

Almost everything that makes Twitter usable was invented by the people using it, and the company has still not worked out how it intends to make money.

The at sign in front of a username was a convention before it was a feature. The hash in front of a word was somebody’s suggestion that stuck. Quoting another person’s message and passing it along started as people typing letters manually in front of copied text, and only later became a thing the service understood. That is an unusual amount of product design to have been done by strangers for free, and it is the single most interesting fact about the company.

What it suggests is that Twitter is closer to a protocol than a product. It does one small thing, badly documented, and the behaviour grows on top of it in ways nobody planned. The fail whale showing up whenever anything happens is almost supporting evidence: the demand keeps arriving faster than the engineering, because the demand is not coming from features they shipped.

The plane in the Hudson in January is when I stopped treating this as a toy. A passenger on a ferry posted a photograph before any newsroom in the country had a camera pointed at the river. That is not a better version of an existing thing, it is a different thing, and every news organisation currently signing up an account is going to spend the next few years discovering how much of their value was distribution rather than reporting.

Then there is the part everyone avoids. There is no business here yet. Not a small one, not a promising one, none. The standard answer is that you build the audience first and monetise later, which worked for exactly one company and is quoted by hundreds who are about to find out it does not generalise. An audience with no revenue is not an asset, it is a running cost, and the bigger it gets the more it costs.

The obvious move is advertising, and advertising sits badly on something this short. A hundred and forty characters has no room for a banner and no patience for an interruption.

My worry is not that Twitter fails. It is that the pressure to find revenue arrives before the culture has set, and the thing that made it worth using turns out to have been the part nobody was charging for.