The last 20% is the product

For years I built B2B software. The last big one was a remote TV management system — started as a box product, ended up white-label that 10 clients could host under their own logo. It made money. Nobody ever complained that the empty states lacked custom illustrations.

Before that, I used whatever tools companies picked. Jira, which I never liked. Trello, which was great for small teams but fell apart the moment you had more projects running in parallel. I'd end up writing my own integrations around whatever was chosen. You moved fast, you plugged the gaps, you shipped.

TL;DRIn B2B you call it done at 90% — in B2C, that remaining 10% is where 80% of the real work lives, and skipping it is the difference between software that works and software people want to use.

B2B is a village of identical houses

In B2B at its cleanest, you're building with LEGO. The same forms, the same validations, the same component patterns over and over. You set up the architecture correctly once — automated tests, form validation, field validation, shared components — and then you assemble. Snap pieces together and move on.

The houses in the village all look the same. That's not a flaw — that's the point. The goal is that it works, that it's consistent, that it scales. Whether a user feels something when they click a button is not part of the spec. B2B users don't care how cool it looks or whether there are animations. They just need to get their work done as fast as possible. The software is part of their workflow, not their experience.

You do 80% of the work, the functionality is there, and you call it done. The remaining 20% — the animations, the error states, the micro-feedback — you skip. Nobody asks for it. Nobody notices.

In B2C, the last 20% is 80% of the work

Then I started building Edicek as a consumer product. First as a web app, then responsive mobile, and now a native iOS app with the first version just approved on the App Store.

I started spending time on things I'd never touched in B2B. Not because something was broken. Because the 90% I would have shipped before wasn't actually done.

Consumer products have a much smaller functionality surface than B2B. Everything that exists needs to be polished completely. You can't just make something simple. You have to make it look simple. Underneath that apparent simplicity, every small interaction needs to be handled — because you're working with someone's subconscious. Every tap, every transition, every error state sends a signal. Does this feel right? Does this feel cheap? Is this worth my time?

Pareto flips. Features take 20% of the effort. The visual feel, the error states, the micro-feedback — that's 80%.

The copy button that vibrates

I spent three days last week on microinteractions in Edicek, on the mobile version.

One of them: when you chat with an AI agent and it finishes its response, a copy button slides up from the bottom of the message. Not before — after, once the answer is complete. You tap it: a small vibration, the icon switches to a checkmark for a moment, then back.

In a B2B tool, I'd leave the button there permanently. Static, always visible, done. The implementation difference is maybe an hour. The feel difference is significant.

There are probably a hundred, maybe two hundred interactions like this in the app. Each one is small. Each takes time I'd never have spent building for a company. Together, they're the product.