Streaming Music Startup Jelli is On a Roll

It’s a Web 2.0 name, but its success hinges on its adoption by traditional media.

Meet Jelli, a San Mateo-based startup that’s aiming to put traditional radio listeners in charge of what songs they hear — and making headway toward that end.

The company, founded in September 2008, already controls the airwave’s of San Francisco’s Live 105 station six nights a week, its users voting songs on and off the air from 8 p.m until midnight. Monday, the service — which has been likened to group-controlled Pandora — launches on 107.5 in Las Vegas, where it will also take over its nightly show. And thanks to a syndication contract Jelli signed with Triton Media in March, Jelli is being pitched to roughly 4,500 other radio affiliates, 17 of which are experimenting with the service now.

So how does it all work, who is using the service, and how big an opportunity is the company chasing? Yesterday, to learn more, I caught up with the company’s CEO  Mike Dougherty.

How did you come up with the idea behind Jelli?

I was [the VP of biz dev] at TellMe in Mountain View, and during the two hours a day I spent on average commuting to work, I started thinking two things: first, about how much of an opportunity the Web in the car will be, and that car radios haven’t seen a lot of innovation in decades. It got [me and cofounder Jateen Parekh] to thinking: large category, great distribution, ubiquitous with established economics — the radio advertising market is $16 billion in the U.S. and $40 billion globally. What if we could develop a Web platform for the radio industry? How might users reengage with a device they all own?

How did you strike on the crowdsourcing piece?

Digg was very popular at the time, and voting is an interesting concept — and one already related to request radio. And we liked the idea of connecting the social Web and connecting it to how broadcasts are created.

How many people have registered for your beta service?

Close to 300,000 people are listening to Jelli, and a portion are signing in and starting to participate but it’s a small enough group that we’re not talking about it yet. It’s worth noting that in the SF market alone, you’re dealing with 5 million people. It’s early. On a nationally syndicated program, you can have tens of millions of listeners.

How do users’ votes get communicated to the radio station?

We created a platform that sits on a server, and the server gets integrated into the radio station tower infrastructure so that when we’re on the air, our server takes over. And our server is constantly connected to the cloud by an application that we developed; it sells the server at the radio station what playlist is being created and rejiggered.

Jelli has a lot of game mechanics built into the service. What can users do right now and how important is that aspect of your business?

Because this isn’t one to one but one to many, we wanted to make the service feel more personalized for the audience, so for example every track has a score and when you vote, you’re either adding a plus one or minus one on the track so you can see how that impacts the score. As its score increases, the track you like starts bubbling up the playlist to the on-deck circle of what comes next. It’s a democratic way of creating order.

Of course, if you’re in a big market like SF or LA, you might wonder: how does my little vote matter? So we decided to give users some limited power ups, like a game. The first is a rocket, where you can apply a rocket on the track and it takes the track from the back of the catalog and shoots it up almost to the very top. Or you can also do a bomb and blow up a song’s score to zero to move it out of the way.

These are pay-for virtual goods?

Not yet. Right now, you have to earn it. If you register and play and vote every day, you can earn rockets and stockpile them and save up to 11 rockets. It’s sort of like the thumbs up and thumbs down of Pandora, except instead of just you dictating, it’s a tug of war with everyone clicking up and down.

It’s hard to believe that’s appealing to a radio station.

The idea was a little controversial when we first started talking about it with radio stations but if enough people click their “sucks” button when something is on the air, it gets blown up midstream, and it’s kind of fun. It sounds like PacMan getting terminated. It doesn’t happen very often during broadcasts but users like that it’s there because it shows that they really are in control. In fact, the first night we were on Live 105, someone rocketed up “Sex on Fire” by Kings of Leon and it was already so overplayed that it got blown up.

You don’t have a mobile app yet, which seems crucial. Is that coming soon?

Mobile was part of our first PowerPoint deck. We’re just now getting around to releasing it. We just didn’t have the resources before with just angel funding. [Jelli raised a $2 million seed round from First Round Capital, TriplePoint Capital and numerous angels in January. A $7 million Series A followed in May.]

Something like 66 percent of radio listening happens outside the home. So mobile is key. The iPHone will be a key part. We’re probably going to be also releasing [an] Android [app] and even looking at SMS. That’s still a very powerful interface.

Who are your users?

Eighty-five percent are 35 and under and 10 percent are true teens.

And how are you marketing the service?

Ironically, we don’t have a robust plan outside of traditional media — radio. For example, in Las Vegas, this week they’re doing pre-promotions on the air. And when we’re on the air in a market, there are calls to action throughout the broadcast — “if you want to join in, log on.” We’ll also try to do fun special events that integrate back to Jelli around holidays or concert series. But we haven’t started a Web-based user acquisition strategy. We’re still trying to make sure user experience is sticky before we scale it.

What’s your pitch to radio stations that might be worried about handing over their airwaves to a Web startup?

That their users are migrating away from them and we can be their bridge back to them.

Is there a financial incentive for them to use you? Could they lose money?

We’re keeping the economics straightforward. In one hour on the radio, there will be a couple of ad blocks that are four to six minutes in length. We effectively pause while those ads are running, and we share in that advertising opportunity. We effectively get paid in similar fashion to syndicated radio shows, like an “American Top 40 with Ryan Seacrest,” except we don’t get paid as much.

Do stations get a free trial period?

No. If we don’t work, we get canceled. From the beginning, we’ve said we wanted compensation from FM broadcast. We’re also monetizing online; our two biggest categories will be ads and virtual goods. But our core revenue will be that FM ad sharing.

You don’t seem to have a huge catalog of songs right now.

It’s growing. Right now, we’re pretty good with alternative rock, but not with hip hop, we pretty much have no country music coverage, and we’re just kind of okay with pop music. With major categories, stations will help us catalyze our catalog growth; otherwise, we’ll launch online-only catalogs.

How does all the licensing work?

For the direct-to-consumer, online-only streams, we’re licensed like Pandora. On those channels where we are partnering with a radio station, the catalogs are tuned by the stations around their format and the licensing is covered by those stations. They pay it based on the licensing that they already have in place for their radio station. So it’s much more of a hybrid model than a pure-play consumer model over the Web.

And what’s the revenue split between Jelli, the station, and Tritan?

Of all the economics, the first cut is what an affiliate pays for Jelli. After that’s calculated, Tritan and Jelli split the rest. Tritan is one of four large syndicators in the U.S.; they have a dozen salespeople and they are our channel to out and pitch these stations, so…

But profitability is imminently achievable. The radio business is so big that even in a recession, the L.A. [radio ad market] is $500 million to $700 million alone. There’s real revenue here.