What should a small business automate first?
Whatever is repeated every day, follows the same steps every time, and costs you something when nobody gets to it. For most businesses that is answering customer messages, because the cost of missing one is a booking that went elsewhere. Do not start with the most annoying task. Start with the most repeated one.
Last checked 14 August 2026. Prices on this page are ours and current.
Three questions that find it
How often does this happen. Daily is worth automating. Monthly almost never is, no matter how much everyone hates it.
Does it follow the same steps every time. A task with a decision in the middle that depends on judgement is a bad first candidate. A task that is look something up, write it down, tell someone, is a good one.
What does it cost when nobody does it. This is the one people skip, and it is the one that decides whether automation makes money. An unanswered message at 9pm can be a booking lost to whoever replied. A pile of unfiled invoices costs you a slow month end and possibly paying something twice.
Where most businesses land
Customer messages, usually. Not because it is the worst job, but because it happens constantly, the answers repeat, and the cost of silence is immediate and invisible. You never find out about the customer who messaged at 10pm and booked somewhere else.
Paperwork comes second, and it is the one that surprises people. Nobody thinks invoice entry is urgent, because nothing breaks when it is late. Then month end arrives and somebody loses two days, and the same somebody is the person who was meant to be chasing payments.
Reporting is third and it is the cheapest win of the three. Most businesses do not need a dashboard. They need one message a day that says how yesterday went.
What not to start with
Anything that happens a few times a year. The payoff is too small and you will have forgotten how it works by the next time.
Anything where the rules are still being argued about internally. Automating a process nobody agrees on just moves the argument into software, and now it costs money to change your mind.
And anything where a mistake is expensive and hard to spot. Start where errors are visible and cheap. Build the habit, then move up.
Start smaller than you want to
The instinct is to fix everything at once, and it is the reason most automation projects quietly die. Pick one task. Run it for a month. See what it actually changed.
That also gives you a real number for the next decision instead of a guess, which is worth more than the time you saved on the first one.
Do not take our word for it.
The easiest one to test in five minutes, because you already have a folder of it somewhere.
- How do I know if a task is worth automating?
- Daily, same steps every time, and something goes wrong when nobody does it. If a task fails all three, leave it alone for now.
- Should we automate before hiring?
- Often yes for the repetitive half of a role, because it is cheaper and instant. It does not replace judgement, and any vendor who says otherwise is selling.
- How long before it pays for itself?
- For customer messages, usually the first booking you would have missed. For paperwork, the first month end that takes hours instead of days. Both are arithmetic you can do yourself before you commit.
- What if we pick the wrong thing?
- Then you have spent one month and a small amount finding out, which is why starting small matters more than starting perfectly.
Tell us what you are trying to work out and we will answer it straight, including when the answer is that you do not need us.
Or email hello@ahlelak.com