What a small business should automate first: examples and a way to decide
Find out which tasks to automate first in your small business: concrete examples, a table to help you decide and the diagnostic I run before building.
Short answer: a small business's first automation should be a task that comes round every week, always follows the same steps, takes up the time of someone who should be doing something else and carries little risk if it goes wrong. In most businesses, that points to follow-ups, paperwork, bookings and the work that comes after meetings. To find out which one is yours, you need to measure before you build.
Why this decision matters
A Factorial study found that in Portugal 46% of managers' time goes on manual tasks that could be automated, and that seven hours a week are lost just checking information across platforms. So it is no surprise that automation is already a priority for 28% of Portuguese SMEs, according to a GfK survey for Cegid.
The trouble is that wanting to automate is one thing and knowing where to start is another. If you start with the tool, you often end up with one more subscription and the same manual work. If you are still unsure about the timing, read When does your business really need to automate first. If you have already decided to go ahead, the next step is choosing the right first task.
How do you choose the first task to automate?
Use four questions and score each one from 1 to 5:
| Criterion | Question | High score when |
|---|---|---|
| Frequency | How many times a week does it happen? | It happens every day |
| Time | How long does it take each time? | It adds up to several hours a week |
| Rules | Does it always follow the same steps? | Anyone could do it with a checklist |
| Risk | What happens if it goes wrong? | The mistake shows up straight away and is quick to fix |
The task with the highest total is your candidate. Notice that the most annoying task does not always win: something everyone complains about might only happen once a month.
The same test tells you what to leave for later. A process that changes every month scores low on rules, because you would be automating a version that will soon no longer exist. And a task that needs a decision case by case can be supported by automation, but not handed over to it.
Which tasks usually come out on top in a small business?
With this table in hand, these are the four that most often end up at the top.
Follow-ups on unanswered quotes. You sent the proposal and the client has not replied. Someone has to remember to get back in touch and, on a busy day, that slips. A workflow that checks pending quotes and prepares the reminder leaves you with nothing more than reviewing it and sending it.
Invoices and paperwork for the accountant. Collecting invoices, naming them, sorting them by month and sending them to the accountant always follows the same steps. It is work with clear rules and high frequency, which makes it a natural candidate.
Bookings and confirmations. If your business runs on bookings, every confirmation by phone or message is repeated dozens of times a week. It is one of the simplest automations to build and one of the quickest to make a noticeable difference.
The work after meetings. Writing up notes, working out what was agreed and creating the tasks is almost always the same. In my business, the transcript of every meeting is analysed automatically and I receive what the client wants and what I need to do next.
Why do I always start with a diagnostic?
The table above works, but it has a limit: the scores come from memory. And memory is misleading, because we remember what annoys us better than what actually takes up our time. Without data, you end up automating what seems to take time rather than what really does.
That is why, before building any automation, I run a diagnostic in four steps:
- In-person meeting. I start by getting to know the company's story and the pain points of the people who work there, because without context the numbers say nothing.
- One week of task logging. The team records what they do on a platform I built for this purpose, which shows where the time really goes.
- Tools audit. I look at what the company already uses, because the solution is often to connect what is already there rather than buy yet another tool.
- Cross-analysis. I bring the three sources together to see where time is being lost and what makes sense to automate first.
At the end, you receive a report with the findings and a presentation meeting, where I explain what I found and in what order it makes sense to move forward.
What happens after the diagnostic?
The diagnostic tells you what to automate, and the size of the job depends on what the data shows:
- One workflow, when one task clearly stands out from the rest.
- A pack of three workflows, when the lost time is spread across several tasks.
- A bespoke system, when you need a solution built from scratch for the way you operate.
Once it is up and running, an automation needs looking after, because tools change and so does the business. That is what the monthly maintenance is for: making sure what was built keeps doing its job.
Where to start
If you do not know what your first automation should be, you do not need to guess. Request the Diagnóstico Flash through the quote form and we will start by finding out where your time is actually going.
On this topic. In a few words.
What is the first thing a small business should automate?
The task that happens most often, with the clearest rules and the lowest risk. In many businesses that means follow-ups, paperwork for the accountant, bookings or the work that comes after meetings.
Do I need to replace the tools I already use?
Not always. Part of the diagnostic is understanding what you already have, because many automations come down to connecting tools the business is already paying for.
How long does the diagnostic take?
Task logging runs for one week. On top of that there is the initial meeting, the analysis and the presentation meeting.
What should I not automate?
Processes that still change often and decisions that depend on human judgement case by case.
