The audit before the automation: why we don’t start with a build

The most common failure mode in automation isn't a bad build. It's a good build of the wrong thing. Someone gets excited about AI, picks a process because it's visible or annoying, and commissions a system for it — without first checking whether that process was ever a good candidate for automation in the first place.

What actually makes a process worth automating

Not every repetitive task is a good target. The ones that are worth it tend to share a few things in common:

• It happens often enough that the time adds up — daily or near-daily, not once a quarter.
• The decision inside it is mostly rule-based, not deep judgment — 'if X then Y,' not 'it depends on reading the room.'
• Getting it wrong occasionally is recoverable, not catastrophic.
• It's currently bottlenecked on a person being available, not on a decision only a person can make.

A process missing most of these isn't a bad idea to eventually improve — it's just not where an automation budget should go first.

Why we scope this before touching a build

An Audit or Systems Architecture Review exists specifically to check this before any code gets written. It's a short, fixed-price look at one process end to end: the tools involved, the handoffs, where time actually goes, and whether automating it would meaningfully help or just move the bottleneck somewhere else.

Sometimes the honest answer is “not yet” — the volume isn't there, or the process is too judgment-heavy to hand off safely right now. That's a real, useful outcome, even though it doesn't lead to a bigger invoice. We'd rather tell a client that early than build something expensive that quietly gets abandoned three months in because it was never the right target.

The quick win, before the bigger build

Most audits end with one small, contained automation included — something that ships in days and proves the model works before anyone commits to a larger system. It's a cheap way to build trust in both directions before either side is asking the other to bet on something bigger.