Guides · Why projects stall

Why do AI pilots never reach production?

Short answer

Because most pilots are designed to demonstrate that a model can do something, not to survive the conditions of the real process. The failures cluster in four places: no agreed success metric, no exception handling, no owner after launch, and integration left until last. All four are decisions made before any code is written.

95%
Of enterprise GenAI pilots showed no P&L impact — MIT Project NANDA, 2025
4
Failure modes that account for almost all of it
Before the build
When each of them is actually decided
Updated August 2026 · Written for people scoping a project, not for search engines

The number, and what it does and does not say

MIT Project NANDA’s 2025 report "The GenAI Divide: State of AI in Business" found that around 95% of enterprise generative-AI pilots produced no measurable effect on the P&L. The figure gets quoted as evidence that the technology does not work. That is not what it measures.

What it measures is a delivery failure. The pilots largely did what they were built to do; they were built to show capability rather than to change a process, and capability does not appear in a P&L. The 5% that did show an effect were, overwhelmingly, projects wired into a real workflow with a number attached to them from the start.

Failure one: no success metric agreed before the build

If the metric is chosen after the pilot runs, the pilot cannot fail — and it therefore cannot succeed either. Everyone declares partial victory, the project quietly ends, and nothing is deployed.

The fix is boring and effective: name one number, measure it before you start, and agree in advance what value counts as success. "Average handling time on this ticket category, currently 11 minutes, target under 7" is a metric. "Improve customer experience" is a way of avoiding one.

We agree the success metric in the audit, before the pilot is scoped. If we cannot find a number both sides will accept, that is a finding in itself, and usually means the process is not the right first project.

Failure two: exceptions were out of scope

Pilots are demonstrated on the clean 80%. Production is mostly the other 20%: the malformed input, the supplier not in the master data, the customer who is also a partner, the case the rules did not anticipate.

A pilot with no exception path looks impressive and cannot be deployed, because the first day in production produces a case it silently mishandles. Designing the failure paths — retry, route to a person, stop the flow — is not polish added at the end. It is the difference between a demo and a system.

Failure three: nobody owns it after launch

An automation is not a deliverable that stays finished. Suppliers change formats, a system releases a new version, volumes shift, a policy changes. Without a named owner and a dashboard that shows drift, the flow degrades quietly and someone eventually turns it off.

This does not require a team. It requires one person whose job description includes it and a monitoring view that makes degradation visible before a customer finds it.

Failure four: integration left until the end

The most expensive sequencing mistake is proving the model first and connecting the systems afterwards. Integration is where the schedule risk actually lives: credentials, sandboxes, rate limits, a system with no API, a vendor who takes three weeks to answer.

Doing it in the other order — establish that the data can move, then prove the model on real data in place — costs nothing extra and removes the failure mode where a successful pilot cannot be deployed for reasons discovered in month three.

A pilot that runs on exported spreadsheets has not been tested against the thing most likely to kill it.

What a pilot designed to survive looks like

One process, end to end, on real data, in the real systems, with a named metric measured before and after, an exception path for every step, and an owner from day one. Narrower than the usual pilot in ambition and far wider in what it proves.

That is the whole argument for scoping tightly: a narrow thing in production beats a broad thing in a slide deck, and it is also the only version that produces the evidence needed to fund the next one.

Follow-up questions

What people ask next.

It comes from MIT Project NANDA’s 2025 study of enterprise GenAI deployments. Like any single study it should be read as an indication rather than a constant, and the important part is not the exact number but the pattern it describes: pilots that were never wired into a process do not show up in financial results.

Still unsure whether your process is worth automating?

Bring us the process. We take it apart with you at no charge and give you a straight answer, including when the answer is no.