I spent most of my life being the youngest kid in the room but now I am the old man in the room. In other words… I’ve been doing this a long time. I’ve watched the same pattern play out across insurance, banking, aviation, manufacturing, telecom, and now AI. And I’m done pretending the traditional discovery phase is anything other than what it is: a very expensive way to avoid building something.
Here’s how it goes. You kick off an engagement. Everyone agrees that before anything gets built, you need to understand the requirements. So you spend weeks, sometimes months, in workshops and interviews. Stakeholders align. Decks get made. A requirements document that nobody will fully read gets produced, reviewed, and signed off on.
Then someone builds a prototype.
And it’s wrong.
Not because anyone is incompetent. Because you cannot discover what a system needs to be by talking about it. You can only discover that by using it. Every requirements document ever written is a best guess from people who haven’t seen the thing yet. By the time you find out which guesses were wrong, those assumptions are already baked into every stakeholder, budget, and timeline in the engagement.
My discovery sessions run about two hours. Then I build. The first working prototype is usually done the same day. Total time from “here’s the problem” to “here’s something you can touch” is eight hours, give or take.
If that prototype is wrong, we just learned something that cost a day instead of three months. If it’s close, we iterate from something real. Either way, we’re moving.
By the time a traditional engagement finishes discovery, I’m usually done with the project.
I just want to enjoy building again
The status quo takes a builder who got into this because they love solving problems and turns them into an accountant. Every late idea from the client becomes a scope negotiation. Every “what if we also…” becomes a conversation nobody wants to have. The consultant stops caring about the best solution because the best solution wasn’t in the SOW. The client stops bringing ideas because they already know how it ends.
Both sides walk away relieved it’s over. A good engagement shouldn’t feel like something you survived.
When implementation time stops being a constraint, the whole dynamic changes. Late requirements are just information. A feature that needs a complete rebuild might add a couple days. A pivot costs what it should cost: the time it takes to move. The client starts saying what they’re actually thinking. The builder gets to solve the real problem instead of the documented one. The work ends up being what both of them wanted, because there was finally room to get there.
The savings are real. Getting the joy back into the process is the part that makes people call again.
The consulting paradigm is long overdue for disruption
The old model, billable discovery phases, multi-month cycles, twelve-person teams as a signal of credibility, is being taken apart. Not gradually. The practitioners who see it are winning at a scale that’s genuinely hard to explain to people still operating the old way. The practitioners defending the old model are not just behind. They are people watching a tsunami get closer and closer to them as they stand on the beach refusing to believe it is anything other than a normal wave. And sadly… most of them don’t even know it… yet.
What changes isn’t just how fast existing work gets done. It changes which work is worth doing at all. Problems that were too niche, too specific, too small to justify a full engagement are now completely viable. You can build something genuinely useful for one person and one workflow without a six-figure budget and a six-month timeline. The level of customization that’s now possible has no real precedent in the industry.
Software can fit the person. Not the other way around.
I’ve been in this long enough to have heard a lot of claims that everything was about to change. Most of them were eventually true and wrong about the timeline. This one isn’t. The technology is ready. The methodology exists. We are here.
The industry has been making this promise since its inception. It kept getting partway there, and then not quite. I think we finally closed the gap.
I’ve been saying this in talks for a while. Figured it was time to write it down.