Is the MVP Still Valid?

If you work at a startup, you can't avoid the word MVP. Over nearly three years at startups, I've built more MVP-sized things than I can count. But at some point, two questions wouldn't leave my head. Is what we're building actually an MVP? And is Eric Ries's The Lean Startup still valid today — especially that famous "MVP"?

Comparing my first day on the job to now, the development environment has completely changed. AI tools have become everyday equipment, and user expectations have risen with them. "Just ship something rough and validate it" used to work; I'm no longer sure it does. So I want to organize my thoughts based on the past few years of experience.

Today's MVP Is Something Different

Back in 2011, the MVP concept was genuinely innovative. Test whether people actually want something with the minimum set of features. In an era when waterfall was more familiar than agile, building anything was expensive and slow, so the approach made a lot of sense.

But things are different now:

• Users expect a certain level of polish from day one. If it feels slapped together, they simply don't use it
• The market is flooded with similar services, and competition is fierce
• Thanks to AI tools, you can build something of decent quality far more easily and quickly than before

Recently, a friend at another startup told me about launching a B2C tool the traditional MVP way. Minimal design, basic features only. Users simply didn't use it. "By the time we improved it, people had already lost interest" — that line stuck with me.

What I took from this is that the approach needs to change. The companies doing well these days build something closer to an MLP (Minimum Lovable Product) than a plain MVP. The features may be limited, but they make sure to include something users will genuinely love.

For example, if you're building a habit tracker, don't just ship a checklist and a chart — add an AI character that cheers users on when they hit their goals. Not essential to the core feature, but something that makes people want to keep coming back.

AI Raised the Day-One Quality Bar

Comparing when I joined to now, things have clearly changed. AI tools have massively expanded what a small team can do.

A solo founder I met recently built a language-learning app entirely alone. Not a simple flashcard prototype — they built the interface with generative UI tools, designed the curriculum with an LLM, and even added speech recognition. All in a few weekends. Beta testers were shocked to learn it was a one-person show.

What cases like this show:

• People who can't code build quite decent apps with no-code plus AI
• Designers ship services by handling development themselves
• A single developer pulls off what used to require a team

Of course, it's not perfect yet. Strange output still happens, and a human still has to review and refine. But it's clear we've moved past "just build something rough" — you can now build something reasonably usable from the start.

Build-Measure-Learn Got Faster

The build-measure-learn cycle at the heart of the lean startup still matters. What's changed is the speed and scale.

Build: With AI helping on coding, design, and content, you can try many versions far faster. What took months now takes days.

Measure: AI helps with user behavior analysis and feedback collection too, so you can process more data more quickly.

Learn: This is where humans still matter. AI can find patterns, but interpreting what they actually mean is still on us.

The most interesting change is that you can now test multiple things at once. It used to be: build A, test it, and if it fails, build B… Now you can build A, B, and C simultaneously and compare.

How We Talk to Customers Changed Too

"Talk to your customers" — the lean startup's basic principle — still holds. But the method has changed.

Customer development used to mean interviewing a few dozen customers; now AI lets you stay in continuous contact with far more users. It can't fully replace direct conversation, but as a complement it's genuinely useful.

There's a caveat, though. If you only talk to AI, you can lose the voice of real users. AI always answers positively and supportively; real users are far colder and more blunt.

So What Actually Matters?

In an era where AI makes everything easy to build, what's the real competitive edge? This is what I think about most these days.

What I believe matters:

Data and learning: Systems that get better every time a user uses them. Accumulating proprietary data others can't easily replicate.

Fast adaptation: How quickly you respond to market shifts and user feedback. By the time competitors copy you, you've already moved a few steps ahead.

Personalization: Experiences tailored to each user — the kind that get harder to leave the longer you use them.

None of this is a definitive answer. Technology moves so fast that I don't know if today's thinking will hold in a few years. But at least for now, I believe this is the right direction.

What Still Hasn't Changed

Amid all this change, some things stay the same.

Forming and testing hypotheses: No matter how good AI gets, confirming you're building what users actually want still matters.

Talking to real users: However sophisticated AI simulation becomes, it can't replace talking directly with real users.

A clear vision: As technology advances, fundamental questions — "why are we building this," "what value do we want to give users" — only become more important.

Wrapping Up

After nearly three years at startups, my conclusion is that the core philosophy of the lean startup still holds. It's the concrete execution that has changed a lot.

Is the traditional MVP dead? Not entirely — but it has definitely changed. The bar for "minimum" has risen. What used to be considered premium is now table stakes.

Nobody knows what comes next. Technology is moving too fast. What's certain is that things will keep changing, and we'll have to keep learning and adapting along with them.

We live in an era without clear answers. That makes it more fun, and sometimes more anxious. But isn't that the appeal of startups?

---

This post is a collection of personal experiences and thoughts. If you have different opinions or experiences, please share them anytime.