The Changing Economics of Digital Product Development

At Matters, we believe in learning before building—a principle grounded in the work of visionary thinkers like Clayton Christensen (Jobs to Be Done), Eric Ries (Lean Startup), Alberto Savoia (Pretotyping), and Steve Blank (Customer Development). Why? Because:

• Building is (usually) expensive, and

• Ideas need real-world feedback to evolve, which often leads to meaningful pivots.

But something big is shifting.

With the rise of large language models (LLMs) and generative AI, the cost—especially in time—of creating a working prototype has dropped dramatically. What once required a team of engineers and a few weeks of development can now be achieved by a small, AI-fluent team in a matter of days, or even hours.

Seasoned engineers at Savas Labs and builders we’ve spoken with at top Bay Area accelerators report writing as little as 10% of the code they ship—down from 90% just months ago. Tools like Cursor, Windsurf, and V0.dev are ushering in a new era of “prompt-driven engineering,” or "vibe coding" (see Y Combinator's surveying in early 2025) making it possible for those with strong product instincts (even without formal CS backgrounds) to bring ideas to life.

This shift is more than a speed boost—it’s a change in what it means to build. Less time coding means more time focusing on the human, emotional, and relational dimensions of product design—the things that truly determine whether someone will love what you’re making.

So yes, your MVP still needs judgment and intentionality. But the barriers to entry are lower than ever. If you’ve ever felt the urge to build, explore, or experiment—there’s never been a better time.

We’re excited to ship more, learn faster, and keep asking what Matters.