The 20-Minute Audit That Finds Your First Automation

Most advice about business automation starts in the wrong place. It hands you a list — automate your invoicing, automate your emails, automate your social posts — and leaves you to guess which one matters for your business.

The trouble is that the right answer is different for a dental practice, a building firm and an online shop. A list can’t know that. You can.

This article isn’t a list. It’s a method: a short, structured look at your own week that ends with a single, specific answer to the question what should I automate first? It takes about twenty minutes, needs nothing but a pen, and you can do it today.

Why the list approach fails

Automation only pays off when it removes work that is genuinely repetitive, genuinely frequent, and genuinely follows rules. Miss any one of those three and you get an expensive system nobody uses.

Take invoicing. For a firm sending four invoices a month, automating it saves perhaps ten minutes. For one sending four hundred, it changes the business. Same task, completely different verdict — and no generic list can tell them apart.

So instead of starting with what can be automated, start with what’s actually costing you.

Step 1: Track the repeats (10 minutes)

Think back over the last five working days. Write down every task you or your team did more than three times.

Be specific. Not “emails” but “replied to someone asking about opening hours”. Not “admin” but “copied a form submission into the spreadsheet”. The specificity matters — it’s what turns a vague frustration into something buildable.

A typical list looks something like this:

Answered “how much does it cost?” — 11 times Copied enquiry details from email into the customer list — 9 times Sent an appointment reminder — 14 times Chased an unpaid invoice — 5 times Answered “are you open on Saturday?” — 7 times

Most people are surprised by their own list. The tasks that eat the most time are rarely the ones that feel biggest.

Step 2: Score each one (5 minutes)

Next to each item, note two numbers.

Frequency — how many times in the five days.

Minutes each — how long one instance genuinely takes, including the interruption. If a phone enquiry takes two minutes to answer but pulls you out of a job for ten, write ten. That cost is real.

Multiply them. That’s your weekly cost in minutes for each task.

The list above works out roughly like this:

Task Times Minutes each Weekly cost “How much does it cost?” 11 6 66 min Copy enquiry into customer list 9 4 36 min Appointment reminders 14 2 28 min Chasing invoices 5 8 40 min “Are you open Saturday?” 7 3 21 min

Nearly three hours a week, on five tasks, none of which felt significant on its own.

Step 3: Apply the rules test

Now the part most people skip — and the reason many automation projects quietly fail.

For each task, ask: could I write down the rules for doing this so clearly that a new starter could follow them without asking me anything?

If yes, it’s a strong automation candidate.

If the honest answer is “it depends” — on the client, on your mood, on things you’d have to explain in the moment — then the process isn’t ready. Automating an inconsistent process doesn’t fix it. It just makes the inconsistency happen faster and at scale.

This is usually where the top item on the list changes places with the second.

Step 4: Pick one

Choose the task with the highest weekly cost that also passes the rules test. Not the most interesting one, not the most impressive-sounding one. The most expensive one you can clearly define.

That’s your first automation.

Resist the urge to do three at once. A single automation you can measure teaches you more than a stack you can’t. Once it’s running, you’ll know something you didn’t know before: how much time it actually gave back, and whether the estimate was right. That knowledge makes every subsequent decision better.

What “automating it” actually means

Once you’ve picked your task, it usually falls into one of three shapes.

A connection between tools. The information already exists somewhere; someone is moving it by hand. A form submission that needs to reach your customer list. An order that needs to appear in your accounts. Nothing needs to be decided — it just needs to travel. These are the cheapest to build and the most reliably valuable.

A repeated answer. Someone asks a question you’ve answered a hundred times. Pricing, opening hours, whether you cover their postcode. The answer exists in your head or your documents; what’s missing is a way for it to reach people without going through you. This is where an assistant trained on your own content earns its keep.

A reliable prompt. The task itself is small, but remembering it isn’t. Reminders, follow-ups, the check-in a week after a job finishes. What’s failing isn’t capability, it’s consistency — and consistency is exactly what a machine is good at.

Knowing which shape you’re dealing with tells you roughly what building it involves, and stops you buying a complex tool for a simple problem.

Two things worth being honest about

Automation doesn’t fix a broken process. If three people in your business do the same job three different ways, sort that out first. Otherwise you’re paying to make the mess run automatically.

Some things shouldn’t be automated. A complaint from an unhappy customer. A negotiation. Anything where the person on the other end would feel worse for having spoken to a machine. The goal is to clear the repetitive work so you have more attention for the moments that genuinely need you — not to remove yourself from them.

Do the audit

Twenty minutes, a pen, and last week’s honest recollection. At the end you’ll have something more useful than any list: a specific, measured, defensible answer to what’s costing you most and what to fix first.

Most businesses find between two and four hours a week hiding in tasks they’d stopped noticing. Whether you build the fix yourself, use an off-the-shelf tool, or bring someone in, you’ll be starting from evidence rather than a guess.

And if you’d like a second pair of eyes on what you find, tell us what came out of your audit — we’ll tell you honestly whether it’s worth automating, and what it would take.