Skip to main content

Private beta with 8 restaurants: $0 and a 48h commitment

RestaPro Team
Product & Platform

In January we went from three friends to eight restaurants. The selection was not random: we wanted variety. A Castilian roast house, two neighborhood cafés, a group with two Mediterranean kitchens, a burger joint, a Japanese place, and a bar with heavy walk-up traffic. Deliberately different profiles. If it worked for all eight, it worked for almost anyone.

The deal was clear and in writing: you pay €0 during the beta. In return, you tell us what's broken, and we fix it within 48 hours or explain why we can't.

Why 48 hours and not a week​

One of the most common complaints with management software is that you report a bug and the answer is "we'll add it to the backlog". The backlog, in many cases, is where bugs go to die. We wanted to break that pattern from the start, even if it was uncomfortable in the short term.

48 hours was the right amount of time for the restaurant owner to feel they mattered, but enough for us to avoid rushing into a patch worse than the original bug. If something could not be fixed in that window (because it required a migration, or we needed feedback from more users), we replied with a plan and a date. No deadline, no promise.

What they asked for and we shipped​

A sample of concrete things from those two months:

  • Waitlist. When a day is full, being able to note down a guest and call them if a table opens up. We added it in a week.
  • Quick slot blocking. For days when a private wedding lands and you need to close external reservations with one click. Used to be a three-step process; we got it down to one.
  • Internal notes on the reservation. "Boss's customer", "celery allergy", "bringing a dog". Visible to the floor team but not the guest. A small feature that changed their daily routine a lot.
  • Email templates. They wanted to tweak confirmation copy without asking support. We shipped a simple editor.

What we discovered without being asked​

Real usage data tells different stories than users do. Looking at the logs from those two months, we found:

  • The service screen was opened on mobile far more than on desktop. It changed how we prioritized responsive work.
  • Almost no one used the floor-plan view at peak hours. They preferred the list view. The floor view was for prep, not for execution.
  • Customer searches by phone were three times as frequent as by name. We tuned indexes and autocomplete accordingly.

What we learned from giving the product away​

Giving things away creates an odd expectation. Half of the beta users were happy to contribute; the other half treated it as a low-commitment trial. For launch, we decided that public pricing had to be clear from minute one, with no eternal free trial and no hidden fees. A short up-front conversation about money saves a lot of misunderstanding later.

By the end of February, with the beta closed, only one thing was left: two frantic weeks to open the doors to the public. That is what the next post is about.

ESENCA