I agree wholeheartedly with the call for modularity and slower, more deliberate decisions. But let’s go further, not just to slow down decisions, but to make them reversible. By shipping smaller, learning faster, and designing for change, you turn fat tails into manageable(r) bumps. Ironically embracing reversibility means you need less governance, not more. No more endless committees approving five-year roadmaps. Just clear guardrails, fast feedback, and the option to course-correct.
But looking from an SI PoV if you start pushing modular, iterative approaches, you’re essentially telling clients, “Hey, you don’t need us for 5 years—just for 6 months at a time.” And you can’t hide behind big budgets and long timelines. You’ll have to deliver, fast and often. No more blaming “complexity” or “changing requirements” for why nothing works. That’s terrifying if your business model is dependency, not value. But here’s the kicker: If you don’t do it, someone else will. The more tailored the system, the higher the chance of scope creep, integration issues, but also new Change Requests and Budget Increases. Unfortunately missed deadlines, blown budgets, and broken promises turn clients into enemies. They’ll badmouth you, sue you, or just never call again. So, what’s riskier: Changing, or being the last dinosaur standing?
