But I have spent the better part of my career walking into situations where someone, somewhere, genuinely believed it did.

Some of the best talent I have worked with was nominated to a CI project because nobody thought it mattered enough to send someone important.

It teaches an entire organization not to trust the next one either. We will sit in a room and nod along to a three-year roadmap full of assumptions nobody can verify.

A number that doesn’t add up. A decision that has quietly been wrong for two quarters. A question nobody wants to ask because of where it might lead.

The first waits for a brief, executes against it, and hopes the brief was right. The second goes upstream, confident enough to advise on the things only that function would know to raise.

Most deployments start with a wave of it. Hundreds of people trained, two days of theory, a certificate at the end. Months later much of it has quietly faded, because the training was never anchored to anything the business had actually decided to do.

Most continuous improvement deployments look healthy in year one. Year two is when something quieter happens — and almost no one names it. On the failure mode that is set not at implementation, but at the start.