Philosophy · 3 min read

Designing automation for humans, not against them

Every time I scope an automation project, I start with the same question: what does this give back to the person? Not the dashboard, not the cost line — the person.

Automation has a bad reputation in support because too often it's pointed at people instead of pointed at their busywork. A bot that deflects becomes a wall. A macro that's too rigid becomes a script people resent reading. The technology isn't the problem; the intent is.

The goal of automation isn't fewer people. It's freer people.

Three rules I keep coming back to

• Automate the dread, not the judgment. Repetitive lookups, status updates, copy-paste — automate all of it. The moment a decision needs empathy or context, hand it back to a human, fast and with everything they need already in front of them.
• Make the handoff invisible. The customer should never feel the seam between automated and human. The agent should never start from zero.
• Measure the human outcome. Handling time matters, but so does whether the agent had a good day. If a "win" burns out the team, it isn't one.

Where it starts

The best automation I've shipped didn't start in a planning doc. It started at a desk, with a ticket I personally hated answering — and the quiet thought, nobody should have to do this by hand.

That's still the whole job.