About
I've made the same bet four times
Each time it looked early, sometimes contrarian. Each time it was the same underlying wager: that the leverage in this work was moving, and that it was cheaper to move with it than to defend the place I'd already learned. Here's the arc, because it explains how I work more honestly than a list of certifications would.
Bet 1 · 2010
I bet on the platform.
I started on Salesforce when Lightning didn't exist, Apex was on API version 18, and we deployed by hand-merging change sets through the UI. “Real” engineers built bespoke systems; a CRM platform was where you went if you couldn't. I took the opposite view — that owning a fast-moving platform deeply would compound harder than owning a language. Fifteen years, ten-plus orgs, and three platform shifts later, it has.
Bet 2 · then
I bet on the seams.
Most Salesforce careers go deep on the org in isolation. I went the other way and specialised in how Salesforce connects to everything else — REST and platform-event integrations, middleware, the messy boundaries where systems actually meet. The value in enterprise work was never the single org; it was the estate. Betting on the seams is why the hardest engagements — government, banking, ISV packaging — became the ones I did most.
Bet 3 · then
I bet on myself.
I went independent on the wager that one senior operator, accountable end to end, could out-deliver an agency's rotating team of juniors learning on the client's org. No layers, no margin stack, no handoffs — just the architect who also does the work. It's a narrower business by design. It's also a better product.
Bet 4 · now
I'm betting on the tooling.
AI-native delivery is the same bet again: the leverage is moving, and defending the old model costs more than moving with it. I use Claude Code, Cursor, and local models as a production engine and stay the architect who reviews, edits, and ships. A build a traditional consultancy quotes at two weeks lands in days, at the same quality. Most of the industry is still billing by the hour for work that no longer takes hours. That gap is the whole opportunity.
The through-line
The technology in every one of these bets has already been half-replaced. Visualforce, Aura, LWC, OmniStudio, now AI — I've forgotten and relearned the stack three times. What carried across wasn't any tool. It was a small set of habits: write less code than you think you need, optimise for whoever inherits the org after you, trust the audit over the documentation, and treat writing as the highest-leverage thing you can do. The bets change shape; the judgment underneath them is the constant — and it's the actual thing I'm hired for.
If there's a fifth bet, I don't know its name yet. But it'll be the same one: follow the leverage, move early, and keep the work honest enough that the next person can build on it.