Startup Stall | Swivel Chronicles P2
We were the Wright Brothers, but we refused to fly anything smaller than a 747.
If you haven’t read my last post, it’s a better place to start.
So here’s the plan
We’re a month in. Noah has picked up Figma and I’m getting the hang of React Native. We’ve got our MVP planned out and we officially have an idea of what we’ll need to launch.
The app will go like this:
- Minimal signup flow
- Home page has a countdown to next event, user can enroll by pushing button
- When the event starts, you’re thrown into a video call with a stranger. We named these calls a Vibe Check™. The call has a 3 minute countdown timer
- “How was the vibe?” (basically swipe right or left)
- Repeat 3 & 4
- If you’re lucky and charming, you’ve landed some dates!
Pretty much Tinder and Omegle had a baby.
And here’s the napkin math
You’re swiping through Tinder at 5 SPM (swipes per minute). In 30 minutes, you’ve swiped 150 times. Of those, you’ll get what, 10 matches? Next you’ll spend 3 minutes per match trying to stand out with some quirky overused pickup line. 2 of those will actually lead to a conversation. That pointless conversation will go on for god knows how long. If you’re lucky, you’ll have a date set up for later that week. There’s a 50% chance the date actually happens. Say it does happen though, there’s a 10% chance it leads to a second date?
And here’s the kicker: what are the odds that you know in the first few minutes whether date 1 should lead to a date 2? Our theory: quite high.
So yeah 15 swipes might seem like a better use of time than 1 Vibe Check, but it’s not.
And that’s it! That’s the math we were all in on. What we threw our lives to side for. What we planned to spend the next 5 years proving.
That won’t work because…
Starting anything new, anything daring, you can expect to hear many different reasons why it won’t work. It comes out as something like “That won’t work because X”. Now, as a founder your job is to first thank them for their time and input. Next, you need to make a choice: decide whether their statement is adding wood or water to your fire.
For us, while we acknowledged the challenges people proposed, our fire was only getting bigger. We were a roaring wildfire burning through all habitats in site.
A common one we heard came from people that had been on Omegle at least once…
Jacking on Omegle
We actually have a page in our notes with this title. It’s apparently a common thing for guys to just sit on video calls doing just that. Why? Idk, but here’s a link from our notes for the curious minded.
Having been on Omegle, we couldn’t imagine a worse experience than downloading our app, signing up for an event, and getting matched with some dude with his tool out. And 3 minutes later being asked, “So how was the vibe?” Like no, f**k off.
But we had a master plan to address this issue. In fact, the master plan addressed many of our expected issues.
The master plan
The first 30 users are friends of Noah and I’s. These are 30 cool people. We know that because, well, they’re our friends. None of them will take their tool out.
They get through our first event. If we did well, half of them want to join the next event. They’ll be asked to bring a friend. Their friends sign up with their personal code, and now the friends are linked.
These people will also not take their tool out. Why? Because what friend would show up to a party as a guest of another friend and take their tool out?
Repeat repeat repeat.
“That won’t work because people will take their tools out”. Ha, check!
The perfect algorithm
As I explained earlier, we’re leaning on the idea that 3 minutes of actual face-time with one person is much better than swiping through many profiles.
We weren’t surprised when people suggested “That won’t work because… I don’t want to spend 3 long minutes talking to someone not on my level”.
I’ll use the term “level” to refer to someones overall date-ability. Excuse my shallowness as I try to represent this idea.
So this goes back to our master plan. We’ve now built this graph of people. Each person has:
1. person that referred them
2. people they’ve speed dated
3. people they’ve referred
They start with a score based on 1, person that referred them. The idea is that someone starts near the level of their close friend. Their level changes based on what people they’ve speed dated thought of them. People they refer also may hint at someone’s level.
This was the gist of it. What we dreamed up was actually more complicated, ie. you might get along better with your friends friends better than someone on the other side of the graph.
Hold up, aren’t we supposed to be trying out an MVP?
How did we just go from trying to get our first MVP out the door to all of these bigger plans?
Well… because we’re officially in startup stall mode. Remember in Part 1 when I mentioned the struggles of handing an MVP over to users? Well here we are, 10 weeks in the depths our cave, justifying to ourselves that we just need a couple more weeks before we can go face the light.
But hey, on the bright side, the app is now 70% of the way to being deemed MVP-ready. We’re planning to head to our launch site in a few weeks, London, where we just signed a 6 month lease. What’s a few more weeks of product R&D to get ready to hit London turf running?
Next up
London! My first time outside the country. We’ll be ready to start our official first launch as soon as we land.
Update: Part 3 is live here: https://medium.com/@nickobrien/can-this-bird-fly-swivel-chronicles-p3-20290b1b8f72
