The fresh start hypothesis.
I have a hypothesis that I cannot prove yet. It is based on two observations from my projects:
- From what I have seen, spec-driven agentic development reduces time-to-market by 10–30%. I have also seen claims of more than 100% speedup, but I think they are doing the math wrong 😁
- I have found this approach difficult to introduce into existing products. The larger and older the product, the harder it gets. For some classes of products, adopting it may not make economic sense at all.
A difference of 10–30% may seem small. But I think that over several years, a team with this advantage will have time to run more feedback cycles with users and build a qualitatively different product.
This leads to my hypothesis: in 5–10 years, products whose development process remains unchanged will be pushed out of the market by products being built now or in the next few years.
What do you think?