Moonxi
Book a diagnostic
← Blog Operations

The pilot worked. So why did it never reach production?

Four questions that separate the pilot that dies from the system that runs. Answer them before you spend one more real.

Moonxi · July 2026 · 3 min read

The pilot ran, the team applauded, and six months later it is still just a pilot. The problem is almost never the model. It is that a pilot and production answer different questions — and most pilots were never built to answer the production ones.

A pilot with no owner, no cost and no system is not a pilot. It is a demo.

The four questions

01.

Does the data the pilot used exist every day? Most of them run on a spreadsheet somebody cleaned by hand the night before. In production, the data arrives dirty, late and from five systems.

02.

Who answers when it is wrong? A pilot gets it wrong and becomes an in-house joke. Production gets it wrong and somebody has to know why, how long it took, and what changes so it does not repeat. Does that somebody have a name?

03.

What does it cost per month, really running? A model charges by usage. At pilot volume the bill is irrelevant. At operating volume it decides whether the project pays for itself.

04.

Which system does it live in? If the team has to open a separate tab, nobody opens it. Production happens inside the ERP, the CRM, the support channel — where the work already is.

What to do with the answers

If you answered all four, the road from pilot to production is engineering, not luck. If one is missing, that one is your project — before any new model. In most companies, the missing answer is the first one: the data.

Got a pilot sitting still?

The diagnostic answers all four questions with your data and your systems, in two weeks.

Book a diagnostic →
Read next Your next hire has no salary and no ceiling. → The cloud bill is auditable line by line. Is yours? →