Sarah is the Operations Manager at a 25-person telecoms company. She knows her team is drowning. She knows automation exists. She has even Googled it twice.
But every time she sits down to think about where to start, she ends up with the same problem: a mental list of 40 different things her team does every week, no idea which ones can actually be automated, and a quiet fear that spending money to find out the hard way is worse than doing nothing at all.
So she does nothing. Again.
This is the most expensive decision a growing SMB makes — and it never appears on a balance sheet.
Why "I Don't Know Where to Start" Is a Trap
The hesitation is understandable. Automation has been oversold for years. "Transform your business." "Go fully digital." "AI will do everything." None of that tells Sarah which of her team's 40 weekly tasks is worth touching first — or whether any of them are.
Most advice on this topic makes the problem worse. "Map all your processes first." "Get a consultant in to review everything." "Build a full operational audit before you decide anything." That approach takes months, costs thousands, and produces a 60-page document that sits in a folder and changes nothing.
Here is the uncomfortable truth: you do not need to understand automation to know where it pays. You need three questions.
That is it. Three questions, applied to any process in your business, will tell you whether it belongs on your automation shortlist — or whether you should leave it alone. No technical knowledge required. No consultant needed before you can even begin thinking.
We call this the Repeatable Revenue Audit. It is the same filter we run at the start of every AI Opportunity Audit we deliver, and it takes less than 30 minutes on your first attempt.
Here is how it works.
Question 1: Does It Repeat?
The single most important characteristic of an automatable process is that it happens the same way, more than once.
Not occasionally. Regularly.
If someone on your team does a task — or a version of the same task — three or more times per week, it is a candidate. Volume is where automation finds its leverage. A task that takes 20 minutes and happens once a month is not interesting. A task that takes 20 minutes and happens 15 times a week is costing you £46,875 a year.
Here is the maths: 20 minutes × 15 occurrences × 48 working weeks = 240 hours annually. At a fully-loaded cost of £31.25 per hour (the real cost of a £50,000 employee including tax, benefits, and overhead across 1,600 productive hours), that is £7,500 a year per person doing it. If two people share the task, you are at £15,000. If it has been running for three years unchanged, you have already spent £45,000 on it.
The Repeatable Revenue Audit starts here because repetition is what makes the economics of automation work. You build something once. It runs hundreds of times. The return compounds every week it operates.
The filter: Does this task happen the same way, at least 3 times per week? If yes — it goes on the list.
Question 2: Does It Follow Rules?
Not every repetitive task is automatable. The second filter separates the ones that are from the ones that need a human.
Ask this: if you had to write this task down as a step-by-step instruction list for a new member of staff, could you do it?
If the answer is yes — if the task has a clear logic that doesn't require years of experience, gut feel, or on-the-fly judgement to execute — then a machine can follow those same instructions. Every time. Without getting tired, distracted, or sick.
Think about what this looks like in practice. "When a new lead comes in from the website, add their details to the CRM (the software we use to track client relationships), send them the welcome email, and create a follow-up task for Friday." That is a rule. It has a trigger, a sequence, and a defined outcome. A machine can do that in four seconds with zero human involvement.
Now compare that to: "When a client complaint comes in, read it, use your judgement about the severity, and decide whether to escalate." That requires context, nuance, and experience. That is not an automation target. That is a human task — and it should stay that way.
The distinction matters because a lot of business owners assume automation means removing human judgement entirely. It does not. It means freeing your team from the tasks that require none, so they can focus on the ones that require all of it.
The Dave Test: If you could write it as an instruction list clear enough for a new hire to follow on day one, a machine can follow it too.
The filter: Can this task be written as a clear sequence of steps with a predictable outcome? If yes — it stays on the list.
Question 3: What Breaks When It Fails?
By this point, you have a shortlist. Every process on it is repetitive and rule-based. Now you need to rank them.
The third question does that ranking for you: what is the cost when this task goes wrong?
Manual processes fail. That is not cynicism — it is arithmetic. Humans make errors, especially on repetitive, low-engagement tasks. The question is not whether a process will fail. It is what happens when it does.
Some failures are annoying. A report goes out a day late. A spreadsheet has a rogue duplicate. Manageable.
Some failures are expensive. A quote does not get sent and a prospect goes cold. An invoice goes out with the wrong amount. A follow-up gets missed and a contract renewal slips away. A supplier confirmation gets lost and an installation gets delayed by a week. These are not inconveniences. They are direct revenue leakage — and in a business running on thin margins, they add up fast.
The businesses with the highest ROI from automation are almost always the ones that automate their highest-failure-cost processes first. Not the most complicated ones. Not the most visible ones. The ones where the consequence of a mistake is a client lost, a payment delayed, or a job falling through.
Run back through your shortlist and ask: what is the realistic cost when this breaks? An hour of rework? A week of delay? A lost contract worth £8,000?
Rank your list by failure cost. The process at the top is your Gold Mine.
The filter: What does it cost — in time, money, or client trust — when this goes wrong? High failure cost = high automation priority.
Putting the Repeatable Revenue Audit to Work
Here is what the framework looks like applied to a real scenario.
Sarah's team at the telecoms company has 40 tasks on her mental list. She runs each one through the three questions. Within 30 minutes she has eliminated 34 of them — either they are too infrequent, too judgement-heavy, or the failure cost is low enough to tolerate.
She is left with six. She ranks them by failure cost. At the top: the process for confirming installation bookings with third-party providers. It happens 20 times a week. It follows a clear sequence. And when it fails — when a confirmation gets missed — the result is a delayed installation, an angry client, and a team member spending half a day untangling the mess.
That is her Gold Mine. That is where she starts.
Not because it is the flashiest process. Not because it is the most technically interesting. Because it is repetitive, rule-based, and expensive when it breaks.
The Old Way:
- Manually email each provider to confirm booking details
- Chase unresponded confirmations by phone
- Log outcomes across a shared spreadsheet
- Discover errors the day before the installation when it is too late to fix them
The New Way:
- Trigger automated confirmation messages at booking and 48 hours before installation
- Flag unresponded confirmations automatically for human follow-up
- Log all confirmations to the CRM without manual entry
- Errors surface days in advance, not hours
Result: 6 hours saved per week, near-zero missed confirmations, and a client experience that no longer depends on nobody forgetting.
Summary
You do not need to understand automation to know where it pays.
You need three questions: Does it repeat? Does it follow rules? What breaks when it fails? Run every process in your business through that filter and you will not end up with 40 vague ideas and paralysis. You will end up with one clear starting point — a Gold Mine that pays for itself within months and compounds every week it runs.
The businesses that win with automation are not the ones that spent six months mapping every process. They are the ones that found the Gold Mine and started digging.
Ready to Find Yours?
The Repeatable Revenue Audit takes 30 minutes when you know what you are looking for. If you would rather have us run it on your business — identifying the highest-value automation opportunities and giving you a fixed-price quote before you spend a penny — that is exactly what the AI Opportunity Audit is for.
No obligation. No technical knowledge required. Just a clear picture of where the hidden gold is in your operation.
*Book your free audit here: https://aideal.group/
Thanks for reading!