What to automate first, and what to leave alone

Automation projects don't fail because the technology isn't good enough. They fail because someone picked the wrong task to start with.

The typical bad pick is something impressive that runs once a month and that nobody checks. Six months later, nobody even knows if it's still running.

The three-question filter

Before you automate a task, ask:

How often does it run? Once a month means a small saving, but just as many ways for it to go wrong. Daily or weekly is worth it.

What does it cost if it breaks quietly? This is the one that matters. A broken report build is annoying. A broken lead handoff means nobody called anyone back for two weeks. That second kind needs an alert when it fails, not only a confirmation when it works.

Does it need judgement? If part of the task is someone deciding what's correct, what you need is preparation rather than automation. The machine gathers the data, and a person makes the call.

What's usually worth doing first

Pulling data together for reporting. Not the pretty dashboard, just the part where you no longer copy from three systems by hand. It runs daily, needs no judgement, and if it breaks, you see it straight away.

Lead routing and notifications. Get the form submission to where the work happens, and tell someone. It's simple, and it makes money directly.

Watchdog checks. A daily job that tells you when the conversion count has dropped to zero, when the feed hasn't refreshed, or when a campaign has spent its daily budget by nine in the morning. This kind of automation does nothing but notice. It's often the most profitable kind.

What not to start with

Anything that writes or publishes in your brand's name without a human signing it off. Whether it works isn't the issue. The first time it gets something wrong, it gets it wrong in public.

The metric worth watching

Not how many processes are running, but how many hours you've got back and how many errors never happened. If neither of those improved, the automation is just one more system to maintain.