Before you automate a task, answer these three questions

Most small teams don't have a tool problem. They have a "nobody ever mapped how the work actually flows" problem — and that's exactly where automation goes wrong.

A business buys a shiny AI tool, hands it to the team with no process behind it, and three weeks later everyone's back to doing it the old way, plus a subscription they don't use. I've watched that cycle enough times to know it's not the team's fault. Nobody stopped to answer three questions first.

The order that actually sticks: map, find the bottleneck, pick one tool, train until it sticks.

1. Where does this job actually start, who touches it, and where does it stall?

Not where the software says it starts. Where it really starts — the inbox message, the phone call, the sticky note. Follow the job through every hand it passes through, and mark the exact spot where it waits. That waiting spot is your bottleneck, and it's almost never where people think it is.

2. What does "done well" look like, in plain language?

Not "more efficient." Something you could check: the estimate goes out the same day, the inbox is empty by Friday, nobody has to re-enter the same information twice. If you can't describe the finish line, no tool can find it for you.

3. Who owns the new way once the excitement wears off?

A workflow nobody owns becomes a workflow nobody follows. Name the person who keeps it alive, and train them until the new way is the normal way. If the team can't do it without you in the room, it isn't adopted yet.

The three questions, at a glance.

Answer those three, and the tool decision usually gets easy — often embarrassingly so. Sometimes the answer is one tool for one step. Sometimes it's a simpler process with no new software at all. And sometimes the honest answer is that AI isn't the right fit for this job, and what the business actually needs is process improvement first. That's a good outcome too, and it's one I'll say out loud rather than sell you a tool you don't need.

That's the order that actually sticks: map the work, fix the bottleneck first, then pick the tool — and only the tool that fits that one step. Tool first is expensive. Process first is boring. Boring win

Tool-first vs. process-first — same team, same week, different starting point.

If you're staring at a task that eats your team's week and wondering whether AI fits it, let's talk it through. Book a free fit call and we'll figure out together whether it's an AI job, a process job, or a mix of both — and if we're not the right fit, I'll tell you that too.

— Jake

Founder, Ignite Consulting | AI workflows for small teams, trained until they stick