Daddy, What Did You Do in the Browser Wars? Do you remember the browser wars? Did you make a difference? Do you worry what you’re going to tell the grandchildren when they ask? I have great news, I think you’re going to get a second chance! The belle of 2025, the AI land grab, wants YOU! * OpenAI recently declared aspirations to get tight into your workspace. * Google seeks to keep your data safe, implementing stricter gating targeting “sensitive-scope requests” * Unfortunately this gating triggers on OpenAI’s new tooling resulting in blocking. As the bubble contracts, platforms seek a defensible share for the endgame. Gatekeeping rears its head. Are the big players shifting to moat-building and putting the walls up? This won’t be the last “This app is blocked” you see over the coming months. All of the players will shore up consumer data safety and their own market position resulting in rapidly changing behaviours. What did I learn last time? * Fragmentation and isolationism raise your costs, forcing you to pay twice not for capacity, but simply for parity of capability * Big incentives could be coming for vocal supporters of the one true way, but you wont feel proud taking the win. * Complexity spirals as partners and customer start making choices that affects your ability to technically collaborate fast Do you see this coming? What did you learn last time?

The Moneylender This is the uncomfortable one. The one I learned from that CTO. We did several raises. And under the spotlight, velocity matters more than anything. Imagine a company raising $1m, then $10m, then $100m. Each round meant to last a year. Silly numbers. Doesn’t matter. Stay with me. Now picture this. You’re down to the last $100k of the first raise. 25 working days left. You’re not fixing bugs. You’re not building platform stability. You’re borrowing time. Because in those final 25 days, velocity is everything. And yes, you take the payday loan. Fast. Cheap. Unsafe. Terrible interest rate. You will absolutely pay for this later. But right now? That 25 days might deliver 45 days of output, if you burn it right. It’s frantic. Wasteful. Ugly. And if you land the $10m, you buy the time to clean it up. And if you don’t? You die. No one comes looking for the code. Most leaders don’t talk about it. Many of the best have done it. It’s not reckless. It’s calculated risk, made visible. You cut the deal. Make sure the board agree the terms.

If you’ve ever had to build or evaluate a roadmap that wasn’t feature-led, I’d really like your take. I’ve been in a few conversations recently where people are struggling to justify work that doesn’t ship a feature. Reliability, security, tech debt paydown, developer experience - the value is obvious right up to the moment where the same effort could tangible impact the bottom line with a new capability. I’ve learnt a few things the hard way about making these cases land, and I’m planning to share some of that over the next few weeks. But before I do, I’d rather listen. Cheat codes? Value calculations? Exchange rates? Explicit capacity reservation? What have you seen that actually changed the conversation and led to better decisions?

Is it safe? I once asked a team to ship something they were ashamed of. Rough code. Unmaintainable. Zero error handling. They looked at me like I’d lost my mind. The thing is, every person has a default delivery mode. Some sprint. Some play it safe. Some squeeze the budget. If you want them to work outside that default, you have to give them three things: Clarity - why this piece of work sits here on the iron triangle. Safety - knowing the compromises are deliberate, recognised, and, if worth iI, paid back. Honour - the right to be proud of work that looks wrong, because it’s right for now. And that safety isn’t just inside the team. All stakeholders outside the team needs to know what they’re getting: the speed, the quality, the cost. No surprises. No one made to look foolish for backing it. Without that? They’ll stick to their default. Every time. Teach it. Fix it. Relentlessly explain why. Even codify it in your design notes, so an engineer can point and say: > “This mess? This was the plan.” When that happens You stop getting “default mode” delivery. You start getting exactly what you asked for.

Scaling up should mean stabilising. Maturing. That’s what I thought too. But that calm? It slows velocity. You started with the basics: * Core systems? Solid. * Growing niggles? Gone. * Team? Calm. * Scaling? Prepped. You look like a Safe Bet. Investors start watching. The board gets twitchy. They see what they fear. They look at you: “Are they right for the next stage?” You wanted to be steady. You look stuck. The sharpest scale-ups don’t freeze the model. They flex the iron triangle—fast, cheap, safe— not once, but on every release. They: * Speculate with safety * Invest for stability * Borrow for speed One CTO shipped services at wildly different maturities. Side by side. Constantly. It drove me mad. It looked like chaos. It was beautiful. I’m sharing 3 behaviours that help every engineering dollar go further. Join me. Learn to drive someone mad. In the best way.