Back to Blog

What Killed the Last AI Networking App — And the 3 Things We're Doing Differently

Shapr raised millions and disappeared. Here's what killed it — and the three lessons we built Mētan on.

8 min read

A few years ago, an app called Shapr promised to fix professional networking.

The pitch was simple: Tinder, but for business. Swipe right on people you wanted to meet. Get matched. Have coffee. Build your network.

It made sense on paper. Millions of professionals hate networking. The randomness of it. The awkwardness. The pile of business cards that turns into a pile of guilt. Shapr said: let AI handle it.

They raised money. Got press. Built a user base.

And then they vanished. Gone from app stores. Website dark. No announcement, no pivot, no acqui-hire. Just... gone.

If you search for Shapr today, you'll find a graveyard of “what happened?” Reddit threads and broken download links. The app that was supposed to make networking intelligent quietly became another failed startup story.

We think about Shapr a lot. Not because we're afraid of the same fate. Because they proved something important: the demand is real. People desperately want smarter ways to connect. The execution was wrong.

Here's what we learned from watching them fail.


1. They needed everyone to show up before anyone got value

Shapr's core problem was the cold start. The app only worked if enough people in your city, in your industry, at your experience level were also swiping. If you opened Shapr in Denver and there were twelve people on it, you got twelve mediocre matches and never came back.

This is the classic marketplace trap. You need supply to attract demand and demand to attract supply. Most marketplace startups throw money at the problem — subsidize one side until the other shows up. Shapr tried, but professional networking doesn't have the viral loops of dating apps. Nobody posts “just matched with a great VP of Sales!” on Instagram.

What we do differently: Mētan activates at physical locations. Events, coworking spaces, meetups, conferences. You don't need a critical mass of random users in your city. You need twenty people in the same room. And those twenty people are already there — they came to the event. We didn't build a marketplace that needs to be filled. We built a layer on top of gatherings that already exist.

One profile. Every event. No rebuilding, no re-downloading, no “is anyone else using this?”


2. They stopped at the match

Shapr showed you a face, a job title, and a short bio. Then it said: “You two should meet!”

OK. But why?

“You're both in marketing” isn't a reason to have coffee. It's a LinkedIn search filter. Real connections don't come from shared job functions. They come from shared context. From discovering that you're both trying to solve the same problem, or that you know someone they need to meet, or that you're working on something that fits perfectly with what they're building.

Shapr gave you the “who.” It never gave you the “why.”

What we do differently: Mētan's Connection Intelligence doesn't just match people. It generates a personalized report for every connection — specific business opportunities, personal synergies, learning potential, collaboration ideas. Not “you're both founders.” More like “you're both building developer tools for the European market, they have distribution you lack, and you have the technical infrastructure they need.”

That's not a match. That's a reason to walk across the room right now.


3. They competed with your couch

Shapr was a purely digital experience. Open the app, swipe, maybe message someone, maybe meet up in two weeks. Maybe.

The problem? You were competing with every other thing on their phone. Netflix, Twitter, the group chat, the other dating app they opened by muscle memory. Digital-only networking has the same engagement problem as every other social app: it's optional, it's forgettable, and there's always something easier to do.

What we do differently: Mētan is physical-first. The app comes alive when you're somewhere — an event, a coworking space, a conference. It's not asking you to set aside time to network from your couch. It's enhancing the connections you're already making, in real time, in real space.

When you're standing in a room full of strangers, the motivation to connect is immediate. The problem isn't willpower — it's information. Who here is worth meeting? What would we even talk about? That's what Mētan solves.

Nearby Discovery runs in the background. You walk in, and the people around you start appearing with context. No swiping. No scheduling. No “maybe next Thursday.”

You're already there. We just make sure you don't miss the best conversation in the room.


The real lesson

Shapr didn't fail because AI networking is a bad idea. It failed because it treated networking like a digital product instead of a human behavior.

People don't network from their couch. They network at events, at conferences, in coworking spaces, at the bar after the meetup. The technology should meet them there — not ask them to come to another app, build another profile, and hope enough strangers show up.

The market Shapr tried to build is now worth billions. Event management software is projected to hit $17.3 billion by 2030. Thirty-five percent of meeting professionals plan to use AI matchmaking this year. Even government procurement events are buying AI networking tools.

Everyone knows networking needs to be smarter. The question was never “is there demand?” It was always “how do you build it without repeating the same mistakes?”

We think we found the answer. Not a swiping app. Not a digital business card. Not another badge scanner calling itself AI.

A connection intelligence layer that activates where people already meet, reveals why they should talk, and helps them act before the moment passes.

Shapr proved the world wants this.

We're building what they couldn't.


Mētan provides Connection Intelligence for real-world interactions. One profile across every event. AI that reveals the potential between specific people, not just the overlap between profiles.