Bloomberg’s most valuable application wasn’t proposed by senior management or product managers. It wasn’t part of a strategic plan. It wasn’t conceived as a killer app.

It was pitched and built by engineers working on an unrelated project.

Here’s the backstory.

Bloomberg Message (and its offspring, Instant Bloomberg) is the moat around the company’s business. It allows 330,000 clients in the platform’s gated community to communicate.

In the early 1990s, New York-based engineers were working with colleagues in London on a foreign exchange application. The group set up a bulletin board where people could type in a question and hope someone across the pond would see it and respond.

Aside from not knowing when the message might be seen, the specific question was folded into a general flow of chatter about unrelated topics.

“It was a horrendous way to communicate,” one engineer recounted.

The engineers wrote code to create a database that stored the text and sent out a notification every time someone logged in. That laid the groundwork for one-to-one communication.

The engineers wanted to release the application to clients. They pitched it to Mike Bloomberg who was skeptical. Does anyone need that? Would people use it?

The engineers offered to forgo compensation if they were paid one penny per message. Mike said that if they had that much conviction they should go ahead.

It is not an exaggeration to say that the product they rolled out changed the fortunes of the company and as a result the landscape of global finance.

The function had a revolutionary feature: it was connected to a directory of clients.

The directory meant that anyone with a Bloomberg terminal could look up and contact anyone else. You could bypass the phone switchboard or gatekeepers to reach decision makers directly.

Nothing like that had ever existed. It was the first social network for finance.

Moreover, your message didn’t disappear into the void. You could see if it was opened.

Later, the engineers added a green/red dot to indicate if the person was online and active. The system also allowed senders to retract messages.

It’s hard today to really grasp how far ahead on the product curve those features were.

Gmail wasn’t created until 2004. And only in recent years have platforms such as LinkedIn and Facebook Messenger provided online status, readership indications or an ability to retract.

Bloomberg clients had those capabilities almost three decades ago.

No customers asked for this feature. There was no focus group or UX review, no survey of the total addressable market or prediction of usage. No one said: “Let’s establish baseline metrics to measure what success looks like.”

Engineers saw the need and wrote the code.

In a world that stresses process over product and user counts over usability, it’s a powerful reminder of how leaps in innovation often emerge organically.

(Part of a series of lessons I learned from three decades working at Bloomberg)