Four or five years ago we built and launched a software product called ServiceBoostLinks. It was a do-it-yourself SEO platform for service businesses — contractors, mainly, the trades we work with every day — with a thirty-day free trial and a subscription from about $150 a month. It had its own brand and its own website, separate from Green Thoughts. It had Stripe billing, a reporting dashboard, a library of written and video guidance, and a good deal of backend automation. It ran for six or seven months, peaked at forty-seven subscribers, and we shut it down.
This is written because what those forty-seven people did — and mostly didn't do — taught us more about how contractors actually buy marketing than anything that has worked.
What it was
The idea came from an objection we hear every week. A contractor knows he should be doing something about search. He's quoted six or seven hundred dollars a month for SEO and balks, not because the number is unreasonable but because it's a real line item for a two-truck operation and he can't see what he'd be getting. So he does nothing.
ServiceBoostLinks was meant to be the option between nothing and seven hundred dollars: our process, packaged for him to run himself.
- Reporting — a crude but working dashboard showing where the site stood and what was moving.
- A guidance library — step-by-step instructions, written and some in over-the-shoulder video: how to write better service-page copy, where in the site to make each change, what to fix first.
- Links and citations — low-quality links and directory listings, automated on the back end and sold as an add-on. Described as low quality inside the product, because they were.
- An upgrade ladder — need a new website? want yours audited? want us to host and maintain it? want the done-for-you plan? Each a step up from DIY into the service work.
The pitch was honest. This is the low-grade version. We'll show you how. When you're ready for more, here's the ladder.
What happened
Forty-seven people signed up at the peak. We had the usage data at the time, and it told a clear story.
Roughly two-thirds of them logged in two or three times and never came back. They weren't using it badly; they weren't using it. Another ten to fifteen percent used it extensively, were disappointed when it closed, and were the people the product was built for — although none of them reported a dramatic change in their business, and most had only been on it two or three months, which isn't long enough to see one.
Then there was a small group — four or five — who stayed past the free trial, paid for two or three months, barely logged in, and then came back asking for a refund because they hadn't realized the trial had ended. The trial length was in the email and on the button. One of them, as far as the records showed, had never signed in at all. Most threatened a chargeback. It is a strange thing to pay for something for ninety days without touching it, and it is also, it turns out, a very normal thing.
Meanwhile, acquisition was running on paid ads, which meant the revenue was thinner than the subscriber count suggested, and we had technical issues we never fully got ahead of. Some of the failure was execution, plainly. But by six or seven months, people were dropping faster than they were joining, and we stopped.
Why it didn't take off
The theory had been: contractors won't pay a lot, so give them an inexpensive way to do it themselves. The flaw was in the second half.
Contractors, as a group, are not very comfortable with computers. That isn't a knock — they are extremely good at things most of us can't do — but it means a product whose value depends on the owner sitting down and working through a guide to rewrite his own service pages is a product most of them will never use past the first login. The ones who are comfortable with technology didn't need us: they'd figure it out, or increasingly, ask an AI to do it, or they'd simply pay for the done-for-you service in the first place. The product was aimed at a customer who existed in theory and barely in practice: technically capable enough to do the work, unwilling to pay to have it done, and willing to pay a little to be told how.
There's a second thing underneath that, and it is the uncomfortable one. Price sensitivity in this market often isn't about the arithmetic. We have watched contractors receive six or eight leads a month for jobs worth twenty or thirty thousand dollars and still not want to pay two thousand a month for the leads. The return isn't in question; the willingness is. A cheaper tier doesn't solve that, because the objection was never really the price.
What survived
More than you'd expect from a product that closed.
The people who liked it referred others. Two of them had us rebuild their websites, and both are still hosted with us today. Across referrals and rebuilds, six or seven of our current hosting and maintenance clients trace back to ServiceBoostLinks. The product didn't work as a product. It worked, modestly, as a way to meet people — which says something about where the value in this business actually sits.
The reporting survived in spirit: the dashboard that was crude in the product is now the system behind every client we serve, rebuilt properly and read by us before it's read by the client. The honesty about links turned into not selling them at all. And the segment didn't go away — the budget-conscious contractor still exists and still needs a web presence — but the product for him is now a fixed-scope, low-cost site build with a small monthly plan attached, which captures a slice of the revenue ServiceBoostLinks was reaching for with the one change that matters. We do the work.
The business changed around it, too. Back then we were almost entirely contractors in one or two trades. Today the mix is much broader — lawyers, physicians, CPAs, consultants — and the pattern that emerged from the DIY experiment is part of why. The clients who stay longest are the ones who want the work done well and understand what it's for, and that cuts across trades more than it follows them.
If we built it again
It would be a far better product. The automation is easier now, the reporting already exists, and we understand the audience in a way we didn't then. We'd also almost certainly not build it, because a better DIY tool for people who don't do it themselves is still the wrong tool. The two-thirds who logged in twice weren't waiting for a better dashboard.