For the first few years of running Amazing Design Services, I was the bottleneck. Every enquiry landed in my inbox. Every review request went out by hand. Every monthly report was me, a spreadsheet, and a Friday evening I didn't get back.
That's not a complaint. It's how most small agencies run, and how most small businesses run full stop. The owner is the system. It works right up until you want to grow, take a week off, or start something new, and then it doesn't.
I wanted to do all three. So I made a deliberate decision: anything I did the same way more than a couple of times stopped being a task and became a system.
The first things to go weren't clever. They were the jobs I'd been putting off because they were dull: chasing reviews, following up enquiries that went quiet, sending the same "here's what happens next" email to every new lead, and pulling numbers together for reports.
None of those needed me. They needed doing, consistently, at the right time. That's exactly what a computer is good at and exactly what a busy person is bad at.
Most of it now runs on n8n, an automation tool I host on a small server of my own, tied into email, Google Sheets and a few APIs. An enquiry comes in, gets logged, gets a reply and gets a follow-up if nobody answers. A report builds itself overnight and lands in an inbox in the morning. My call list is waiting for me when I sit down with a coffee, already sorted by who's worth ringing first.
This is the part people skip. I didn't automate conversations. I didn't automate design decisions, or the judgement call on whether a business actually needs a new website or just needs to fix three things on the one they've got. I still pick up the phone.
Automation took away everything around the work so I could spend my time on the work itself. That's the whole point. If you automate the human bits, you end up with a business that's efficient and that nobody wants to deal with.
I automated things before I understood them. If a process is messy when a person does it, it's still messy when a machine does it, just faster and with nobody noticing. Now I do something by hand a few times, write down what "done properly" looks like, and only then build it.
I also learned that automations fail quietly. A person who forgets to send an email usually remembers eventually. A workflow that breaks just stops, and you find out three weeks later. Anything important now tells me when it's run, and shouts when it hasn't.
Without this, there's no 99 Websites and no Review Velocity. I couldn't have started either while I was still the person manually sending every email at Amazing Design Services. Getting myself out of the day-to-day is what made room to build something new.
If you run a small business and this sounds familiar, don't start with a big automation project. Pick the single most boring, repetitive job you do every week, the one you'd happily never do again, and make that one run without you. Then pick the next one.