The thing about building software when you run a service business

Authored by: Callum Gracie

Most operators I talk to want to escape their service business. They hate the people-juggling, the endless Slack pings, the way every Friday becomes a fire drill. So they dream about SaaS. Meanwhile they picture waking up to Stripe notifications while they sleep, sipping coffee in some other country, building once and selling forever. That dream gets sold pretty hard online.

But here’s what nobody mentions. The cemetery of failed SaaS products is full of smart people who had the technical chops and the idea, and zero way to find a customer. Marc Lou put it cleanly in early 2025. SaaS is the worst business to start because you need to build insane trust and solve a painful problem from a standing start. He’s not wrong.

So here’s the counterintuitive part. If you already run a service business, you have the one thing every wantrepreneur is missing.

You already have a distribution moat

Picture the average AI startup launching this year. They have a slick product, three engineers, two million in seed money, and a Notion doc full of beautiful roadmaps. Then they realise they have no audience, no email list, no warm intros, and Google has stopped sending free traffic. Now they spend most of their runway just trying to find someone who will take a demo call.

Meanwhile you, the agency owner with 40 clients, get a reply rate that would make their heads spin. Your clients trust you. They’ve paid you for years. They have problems your product could solve, and you already know the exact language they use to describe those problems.

That’s the moat. Naval Ravikant called it earlier this year. Pure software, he said, is becoming uninvestable. Evan Spiegel said the same thing on Lenny’s podcast. Distribution is the moat now, not features. Features get cloned in a weekend by someone with Cursor and a Red Bull.

So if you’re sitting on a service business, stop hating it. Start seeing it for what it is. It’s a paying audience that funds your next move while you build the thing.

The playbook is older than you think

Look at the canon. 37signals was a web design agency before they built Basecamp inside their own shop. They got 100 paying customers in the first month because they already had a list. ConvertKit started life as Nathan Barry’s freelance design work before becoming Kit. MailerLite was a Lithuanian web agency before it became a 165-person email tool serving customers in 40 countries. BrightLocal was a local SEO agency that turned its internal toolkit into 10 million in recurring revenue and 20,000 customers.

None of these founders quit their service business to build software. Instead they funded the software with the service business. Then they shipped the product to the people they were already serving.

Here’s what makes the difference. Every one of those companies built the SaaS inside the existing distribution. They didn’t try to start a new audience from zero. So when launch day came, they had warm leads, real feedback, and revenue from week one. Most other launches are still trying to figure out the headline a year in.

That’s also why YC’s 2025 request for startups specifically called out service businesses turning into AI-native operators. Not selling tools. Selling outcomes. Owning the workflow.

The trap to avoid

Most operators who try this make the same mistake. They quit the agency too early. Then they spend a year building features in the dark, run out of cash, and end up taking client work again at worse rates than before.

Don’t do that. Run the service business as the engine. Let it pay for the product team. Charge agency clients for early access in beta. Your customers fund your roadmap, and they tell you exactly which features matter because they’re using the thing in front of you. That feedback loop is worth more than any user research budget a VC could write you.

Honestly, this is the part most threads on X miss. The dream is “leave your job and ship a SaaS.” The reality that works is “use your job to ship a SaaS.” Boring. But it ships, and it doesn’t bury you.

What I’d tell my younger self

If I could send one message back, it would be this. Stop chasing the tool and start mining the work. The frameworks you’ve built, the checklists, the weird internal spreadsheet your team uses every day. That’s the product hiding in plain sight. The distribution is already built into your client base. The only thing left is shipping it without burning the engine that keeps you alive.

So pick the niche your service business already serves best. Productize the most repeated thing you do. Charge for the outcome, not the seat. Then keep showing up and let compounding do its job.

That’s the operator move in 2026. Forget founder mode and blitzscaling. Vibe-coded unicorns make great threads on X, but they aren’t the play for someone like you. The real play is patience, distribution, and a service business that earns the right to ship something bigger.

Author Bio:

Callum Gracie, Founder, Otto Media