AUGUST 4, 2026

We Made the Top 10. I Still Didn't Know If We Were Building the Right Thing.

On August 4, 2026, Kizuna Hub reached the Top 10 final round of the Student Innovation & Entrepreneurship Challenge 2026 in Da Nang. I was happy about it. For a project that had spent months living in prototypes, diagrams, and team conversations, an outside signal like that made Kizuna feel much more real.

The strange part came soon after. In a mentoring conversation, we were asked a few questions that sounded almost too basic: who actually uses Kizuna, who influences the decision to adopt it, and who pays? I knew the product well enough to explain almost every screen. My answers became much less clean when the conversation moved to the business around those screens.

1. Reaching the final round made Kizuna feel more real

A competition gives an early product something it usually doesn't have: a moment where the story has to stop moving. You can't keep saying the idea is still evolving. You have to explain what the product is, who it is for, and why anyone should care, to people who don't share any of your context.

That pressure helped us. Kizuna became easier to explain because it had to. Reaching the final round felt like recognition of that progress, and I think the team needed that energy. The mistake would have been letting one kind of progress answer questions the competition never tested.

The mentoring questions weren't about whether the UI looked good or whether the matching flow worked. They were about what a university would need badly enough to pay for, what would make a mentor keep participating, and how we would know that founders were actually getting better outcomes.

There was nothing I could refactor to answer those questions. That was useful in a slightly uncomfortable way. We had spent a lot of time making the product more concrete, while some of the assumptions around the product were still just that: assumptions.

3. We knew our features better than we knew our stakeholders

By then, I could explain the founder workspace, venture context, mentor recommendations, and the flow around a mentoring session without much effort. What I couldn't explain as confidently was why each person around Kizuna would keep coming back. A founder, a mentor, and a university can all be part of the same product and still want completely different things from it.

That also changed how I thought about mentor matching. An introduction isn't the end of the problem. The harder part is whether the mentor has enough context to be useful, whether the founder leaves with a clear next step, and whether that history can help the next conversation instead of disappearing after the call.

4. Who uses it, who influences it, who pays?

The distinction sounds obvious now, which is probably why it was so easy to skip. The person using a product isn't always the person deciding whether it gets bought. We had naturally spent most of our attention on founders because they sit at the center of the experience.

But if a university is the one expected to pay, then 'this helps founders' isn't enough. We need to know what problem the university itself is paying us to solve and what would make that value visible enough to defend the cost. Writing a subscription model into a pitch deck doesn't answer that. Talking to the people involved might.

5. The next version has to begin with evidence, not imagination

The most useful thing about making the Top 10 wasn't that it proved Kizuna was right. It made the gaps easier to see. We need direct insight from universities and mentors about how the current process actually works, where it breaks, and what they would care enough to change.

I still believe Kizuna can become valuable. That belief is enough to keep building, but it isn't evidence by itself. The next version should come from what we learn outside the team, not from getting better at imagining the perfect version inside it.