Clarity on what to build. Honesty about what not to.

I set out to build a pet habit tracker. The premise was the one everybody nods along to: owners log their pet's daily routines, and with enough of that data we tell them something useful.

The assumption I hadn't tested was whether owners would actually track. Not whether tracking is useful — whether they'd do it, repeatedly, unprompted. The entire product sat on top of that, and I'd never checked.

So I checked. Closed-group surveys and two small pilot programmes — a few weeks, unglamorous. The answer was clear: they liked the idea and they didn't do it. The data layer the whole product depended on was never going to exist.

I cut the habit tracker. I cut the AI features stacked on top of it. What survived was the part that was genuinely ours and needed no daily user input — the nutrition analysis engine, built with a certified pet nutrition coach.

That removed roughly 80% of the planned scope and projected build cost, before any of it got written.

It's in alpha, bootstrapped, and approaching its first year. The audience arrived before the product did — 2,500+ organic followers in the pet wellness space, no paid promotion, now generating consistent inbound.

I could have built the original version. It would have been beautifully specified and dead on arrival. The specs would have been excellent. That's the failure mode I exist to prevent, and I know it because I nearly walked into it with my own money.

I started an independent consulting and creative practice for SMEs. Months of research pointed me at well-run local businesses in the US and UK. I built a positioning, built a marketing plan, reached out to owners over email and WhatsApp, warmed them up. Several asked to see a proposal.

I read that as validation. It wasn't. Nobody had paid me anything. "Send me a proposal" is the cheapest thing a business owner can say, and I'd built a company on it.

Then the market fell away.

What I did next was the actual failure. I pivoted toward healthcare, because healthcare was the thing moving. I had no relationships there, no credibility there, and no reason to believe I could win beyond the fact that it was busy. I chased the market instead of asking whether my offer worked. Then I held on for months, because it was mine and I'd built it.

I never landed a client.

What I'd tell myself now:

• Interest is not demand. Get someone to pay before you build a business around them.
• A pivot toward wherever the noise is isn't a pivot. It's panic with a strategy deck.
• The hardest judgment in product isn't what to build. It's admitting the thing you're attached to isn't working, early enough for it to matter.

That last one is why I do this work. I was slow to say stop to myself. I'm not slow to say it to you.

They had a working offline business and no online presence. They came to us for a lead generation website.

I ran the discovery. What surfaced wasn't a demand problem — their tracking and coordination happened offline and manually, and it was costing them money and time on every transaction. Sending more leads into that would have amplified the cost, not the revenue.

So the recommendation wasn't the website they asked for. It was to move the operation itself online — customer, inspector, warehouse, and operations touchpoints, mapped as one connected ecosystem — and let demand generation follow a process that could actually handle it.

Every major functional decision went through their CEO directly. It's live and running. The modules I specced hit 70–80% first-pass UAT acceptance.

The brief was reasonable and the client wasn't wrong to ask. But the thing they wanted to buy wouldn't have fixed the thing that was hurting them. Finding that out took discovery — and discovery is cheap compared to building the wrong thing well.