When Biz Stone came up with the idea for Jelly in 2013, a Q&A app for crowdsourcing answers to everyday questions, he asked around Silicon Valley, looking for someone to dissuade him. Stone was a few years removed from co-founding a massive success (Twitter), and investors were eager to back his next project. He couldn’t find any dissenters.
“I got kinda nauseous, like, ‘Oh my god, I have to start all over again …'” Stone joked in an interview this week with Re/code. “I really was hoping someone would tell me it was stupid and not to do it. People were like, ‘Oh, Biz Stone made Twitter in his garage, and it just works.’ Um, not quite. I did a whole bunch of stuff that didn’t work at all.”
“There’s no correlation between one thing working and the next thing working,” he added.
“You can’t just stay the same. You have to do stuff that might not work. You can’t be too precious about anything these days, you know, because everything moves so fast, and suddenly you’ll find yourself, like, ‘What just happened?'” — Biz Stone
Fast-forward two years, and Stone — who is refreshingly honest for a Silicon Valley entrepreneur — has watched his initial concerns materialize. Jelly didn’t gain traction, so almost a year later, Stone and his team launched Super, a funky social networking app for sharing things with friends. That, too, never took off outside of a small group of dedicated users.
So, earlier this week, Stone announced Jelly 2.0, a return to the team’s original Q&A idea, with a slightly different spin: Instead of posing a question to people you know — a process that can be limiting and possibly embarrassing, depending on the question — Jelly 2.0 will find someone on the app who can answer your question, while simultaneously keeping you anonymous.
“Dear Abby” in the form of a smartphone app.
Stone doesn’t have a product in market yet. He doesn’t even have an official launch date. But he knows what he wants to build, and he’s also aware of what’s at stake.
“Hopefully, the third time’s the charm here,” he said. “I told [co-founder Ben Finkel], ‘We just need to swing for the fences on this one.’ I told Ben, ‘Let’s spend some money. Let’s go crazy on this, because if this doesn’t work, then we’re screwed.'”
Re/code caught up with Stone to find out why he’s reviving Jelly, and what went wrong the first time around. Oh, and we also asked him about Twitter’s plans to expand the length of tweets. (Spoiler: He’s pretty cool with it.) The following transcript has been edited for clarity and brevity:
Re/code: Didn’t you already try this idea with Jelly? What’s different this time around?
Biz Stone: A big thing is we detached identity from queries. Imagine if every Web search you did was also a [public] tweet. That’s what Jelly 1.0 basically was. So, automatically, a huge percentage of what you wanted to ask, you’re not going to ask.
What we’ve also done this time is devise a method for routing queries to people who are more likely to be able to help you. So it’s not just that we’re sending it to your friends; we’re sending it to people who we believe are going to be able to answer your question, because they’ve been there, done that already.
But how do you know what I’m an expert in? Do I have to tell you everything I’ve done?
That’s kind of our computer-science-y stuff. At the most fundamental level, people tell us what they know. I wouldn’t say experts, either. Like I said, it’s people who have been there, done that. On a place where it says, “What are your expertise?” I wouldn’t list “glasses frames in Marin.” But I live in Marin, and go to the LensCrafters in the mall there, and I know what they have.
But that’s my point — you wouldn’t list that. So how do you know that about me?
We can figure that stuff out based on things like related topics. So if you tell us one thing, we now know 10 [things about you]. If you tell us you are into photography, we know that you probably know about lenses and film. The whole service just gets smarter and smarter the more people use it.
So, I imagine as this builds, Jelly 2.0 becomes a search app in addition to a Q&A app?
Fundamentally, Web search indexes documents and matches your queries up with documents that are most relevant. Instead of matching somebody with a relevant document, we’re matching somebody with another person. That’s a living, breathing person who’s had the same light go on in their car, or the same medical procedure, and then you can follow up with them.
So what happens to Super? Will you shut it down?
Right now, it’s happy, it’s thriving, people are signing up every day. There’s a very tight-knit, different kind of community that’s formed on there. It’s a thing that’s growing on its own — that’s hard to do. It’s really not a lot of work to keep it up these days with Amazon Web Services. People are having fun with it, so I don’t see any reason to close it down.
I have to ask — what did you think of the news that Twitter is looking to expand the character limit for tweets?
I love essays. I always thought of articles being like a picture or a video — why can’t you do a tweet, and instead of embedding a picture with it, embed an essay. Everyone does this anyway. They write their Medium post or their blog post or their news article, and then tweet it out. So why not have it be in the Twitter card?
So you’re cool with it?
You can’t just stay the same. You have to do stuff that might not work. You can’t be too precious about anything these days, you know, because everything moves so fast, and suddenly you’ll find yourself, like, “What just happened?” You gotta keep on your toes.
This article originally appeared on Recode.net.