Everyone says "use AI." Almost no one says where to start.
If you run a business, you've heard it a hundred times: you should be using AI and automation. What nobody tells you is where to begin — so most teams either freeze, or they automate the flashiest thing instead of the most valuable one. Both are expensive mistakes.
The good news: choosing your first automation isn't a guessing game. It's a scoring problem.
Don't start with the shiny thing
The most common failure is automating what's visible rather than what's valuable. A chatbot on the homepage looks great in a demo and rarely moves revenue. Meanwhile the rep who spends two hours a day copy-pasting leads between tools is quietly costing you deals — and that's the boring win that compounds.
The second failure is the "big bang": trying to transform an entire department at once. Large projects stall because they touch too many people, too many edge cases, and too many opinions before anything ships. Momentum dies in the planning phase.
A simple test for your first workflow
The best first automation usually checks five boxes:
- High frequency — it happens every day, not once a quarter. Frequency turns small time savings into real money.
- Rule-based — the steps are consistent and explainable. If a competent new hire could follow a checklist to do it, a machine can too.
- Low judgment — it doesn't hinge on taste, negotiation, or reading a room. Keep those with humans for now.
- Measurable — you can name a number that should move: hours saved, response time, error rate, leads followed up.
- Owned by you — you control the process end to end. If shipping it needs three other departments to agree, that's a second project, not a first one.
A quick way to rank candidates: multiply how often it happens × the time it takes each time × the cost of getting it wrong. The highest score is usually your starting point — and it's almost never the thing you'd put in a demo.
What to avoid first
Skip these until you've shipped a win or two:
- High-judgment work — strategy, pricing, sensitive customer conversations.
- Rare, complex tasks — the payoff is too small to justify the edge cases.
- Anything politically owned by another team — you'll spend more time in meetings than building.
- Irreversible, high-blast-radius steps — if a mistake is expensive to undo, that's not where you want to be learning.
Start small, measure, then expand
The teams that win with automation don't start big — they start narrow and measured. Pick one workflow. Instrument it so you can prove the before-and-after. Ship it. Then use that proof — and the trust it earns — to fund the next one.
That's the whole arc: discover the highest-value workflow, design it around your real process, deploy it small, and optimize once the data is in. The first win buys the appetite for the next.
Your first week
If you want to move this week:
- List every task your team does more than once a day.
- Score each one: frequency × time × cost-of-error.
- Circle the highest score that you fully control.
- Write down the single number it should move.
- Automate the smallest version that moves that number.
That's it — not a transformation, a first provable win.
If you'd rather not guess, that's exactly what we do. Book a 30-minute call and we'll help you find the first workflow worth automating.