Skip to content
Building in Public #Founder#SaaS#BuildinginPublic#Consulting#Startups

Six Products in a Year, and the Consulting That Funds Them

· 7 min read

In January I published From Developer to Founder. I had two products, Mr. Popup and Suippy, and the conclusion was that building software is writing code, and building a SaaS is understanding people.

I still think exactly that. Everything else has changed.

Ten months later there are six products. Three charge money. Two are still running. Four are paused, and one of those never launched at all.

This is the follow-up, with figures, because a January article still talking about two products is no use to anyone.

The six

Mr. Popup. Conversion popups generated from your own site, with nothing to configure. Paused.

Suippy. Sell one product online without building a store. Finished, and paused before it launched.

Claim the Valley. An advertising marketplace shaped like a city, where every building is a claimable plot. Paused.

EasyClaw. A managed runtime that put an autonomous AI agent online in under a minute. Past 1,500 users. Paused.

Clinch. Abandoned order recovery for Shopify, measured against a holdout group. Live.

Replyt. Turning Reddit conversations into customers. 94 founders, 3,875 threads reviewed. Live.

The numbers, undressed

Here is what the three that charge have made, all-time:

EasyClaw. US$1,459 total, US$298 in monthly recurring revenue today, across two active subscriptions.

Clinch. MX$5,528 in commissions charged, about US$295. Shopify bills in dollars.

Replyt. US$127 total, US$49 monthly on one active subscription.

A little under US$1,900 in all. That does not support anyone, and it is not meant to. It is evidence that three of six reached the point where somebody takes out a card, which is exactly where most side projects stop.

I am publishing the figures because the alternative is writing about “traction” in the abstract, which is what all of us do when the numbers are small.

The number that does matter

Clinch charged MX$5,528 in commissions. That figure only means anything next to the other one: the stores using it recovered more than MX$190,000 in sales, each tied to a real order ID and measured against a holdout group.

The two figures do not divide into each other, and that is deliberate. I charge no commission in the first month: the store recovers sales from day one and I only start charging the month after. Almost everything Clinch recovers stays with the store, which is how I want it to work.

The product does exactly what I designed it to do. What it does not have yet is scale, and scale is a distribution problem, not a software one.

It is January’s lesson with ten more months on it: building was never the hard part.

The one I paused earns the most

EasyClaw makes US$298 a month. It earns more than the other two combined, and it is one of the ones I paused.

The reason was not the money. EasyClaw solved a real problem: getting an autonomous agent online without Docker, without a server, without DevOps. A few months later there were half a dozen alternatives doing the same thing and the runtime had become a commodity. I could compete on price in a market where price trends to zero, or I could stop.

I stopped. The subscribers who are left still pay and the service is still standing. Paused does not mean switched off. It means I stopped spending time on it.

What pausing four products costs

Here is the part almost nobody writes about, and the reason this article exists.

Pausing a product that bills US$298 a month is an easy decision to make and a hard one to afford. A founder whose income depends on that product does not make it. They stretch it out, add features, cut the price, convince themselves it needs one more push.

I could stop it because my income does not come from there. It comes from consulting.

Consulting is not the fallback

It is the structure.

A three-step cycle. Consulting pays the bills, which lets me build my own products without needing them to pay for themselves, which is why I can pause them when they stop making sense. What I learn building them goes back into the client work. Consulting pays the bills So products don’t have to pay for themselves So I can pause one when it stops making sense And what I learn building them goes straight back into the client work
Neither half works on its own. That is the whole point.

Consulting pays the bills, and that buys something investor money does not buy in the same way: the freedom to kill a product the day it stops making sense, with nobody depending on it staying alive.

It runs both ways. Every product left me something I now get paid for:

Clinch put me deep inside the Shopify App Store, its billing, and an attribution model that has to survive a merchant questioning it.

EasyClaw was AI agent infrastructure in production, built and deployed in twelve hours.

Replyt is distribution, which is still the part I find hardest and the part clients ask about most.

Suippy taught me not to build. It is finished and it never launched, and that decision saves me months on other people’s projects.

When a client asks me whether their idea holds up, I am not answering from theory. I am answering from six attempts of my own, four of which I decided to stop.

What I would tell the person who wrote that January article

Three things.

Building is no longer the bottleneck, and has not been for a while. Six products in a year prove it, and prove it in the bad sense too: building fast only gets you to the hard question sooner, and the hard question is whether anyone wants it.

Pausing is a skill, not a defeat. Four of six are paused and none of those decisions keeps me up at night. Giving EasyClaw another six months rather than accepting what was already obvious is the one that would.

An income that does not depend on the product changes every decision you make about the product. For me that is consulting. It could be a job, or one large client, or anything else. Without it, every product decision carries fear inside it, and fear is a terrible head of product.

What happens next

Clinch and Replyt continue. Both have people paying and both have the same open problem, which is distribution.

Consulting stays at the centre: custom development and technical decisions for companies, which is where my work has been for fifteen years and where I am most useful.

I see no tension between the two. The products are the reason my technical judgement is still judgement and not just memory.


If you are deciding what to build, or whether to build it, that is usually where I come in. Let’s talk, or see what I have worked on.

Related Posts