A couple of blocks away, South by Southwest revelers are already hard at work doing what they've come to Austin to do: Namely, hitting the food trucks, drinking in the sun, listening to good music, and maybe watching a panel or two.

But from where I'm standing on San Jacinto and East Fifth Street, the world looks very different. The year is 1885, according to a yellowed Austin Daily Statesman newspaper in the window of Dandy's old suit shop. Eight young black women have been brutally murdered, and the people of Austin are scared. The killer, known as the "Servant Girl Annihilator," is still crawling the streets. I'm walking to some of the key locations of the crimes to try to unravel the mystery.

At the end of my morbid (but fascinating) tour, I do not actually find out the identity of the country's first known serial killer. Instead, at a nearby bar I find Andrew Mason, co-founder of Detour, the company that created (along with NPR's Radiolab) the audio walk I just took.

If you haven't been paying attention to Mason since he was ousted as Groupon's CEO in 2013, let me get you up to speed. He spent a week at a Miami "fat camp" to slim down, took six months off, had a baby with his wife, pop singer Jenny Gillespie, recorded an album of motivational business songs, and then started this location-aware walking app company.

Mason almost went in another direction entirely. He had been spitballing a different idea--an expert marketplace company, kind of like "an Uber for on-demand expertise," he says. But in the end, he just wasn't passionate about it. And this time around, passion mattered.

The Austin walk, released in time for SxSW, is the latest Detour to come out of Mason's new company. Up until now, the startup has been focused on creating quirky narrated walking tours around San Francisco, where it's based. (One involves investigating the city's trash problems.) New York, Chicago, and L.A. tours will come eventually. A subscription to the app is $20 a year, or you can buy individual Detours for $5.

"It's not that this hasn't been done before," he says of the idea, "but previous attempts were half baked."

So Mason drew on his background in music and his love of travel to start Detour. Traditional audio tours involve staring down at a small screen to make sure you're at the right point in the narrative. He likens the Detour experience to being in a video game, with you as a character. The app uses some fairly sophisticated location technology so that the narrative and music move with you as you walk and pause when you pause, which means you should never need to look at your phone.

But the vision is much bigger than just creating a cool app for travelers; it's to create a new content medium for storytelling. The ultimate goal is to enable anyone to create a Detour with simple, WordPress-like tools. And residents will enjoy them just as much as a city's visitors.

"I think we're building something really beautiful," Mason says. 

In this second venture, an older and wiser Mason is doing things in a decidedly un-Groupon-like fashion. For one thing, he has funded Detour himself so far. He's chucked the notion that startups need to go from launch to product in a matter of months. "We took one year to launch, we spent three months creating our first Detour, and we spend tens of thousands of dollars on each one," Mason says. "We're making a long-term bet by showing what 'good' looks like."

Mason now also has a considerable disdain for that oft-heard Silicon Valley idea that startups should be driven by data. "You have to be religious about certain values and believe them regardless of what the data show," he says. That's the only way to really build great products and a customer-obsessed business. He cites as a counterexample Groupon's decision to A/B test sending two daily promotional emails to customers in Europe instead of one, which Mason says went against all of his customer instincts. Unsubscribes went up, but so did revenue--until it didn't, and the whole market stagnated. 

When it comes to building a team, he now looks specifically for qualities like self-discipline and relentlessness. "In a startup, employees are constantly reaching the limits of their capacity," Mason says. When that happens, do they blame the world for their failure, or seek to grow themselves?

To identify people who choose the latter course, in the interviewing process he sends candidates a series of questions ahead of time so that they can think about their answers and then write them down. People aren't always at their most thoughtful when they're thinking on the fly, after all. "There's no benefit to surprising someone," he says. He fires faster now, too, because he says he's seen how quickly employees who are a poor fit can poison a culture.

His Detour team in San Francisco consists of only 10 people currently, but he's very clear about the fact that he thinks it can be a big company that lasts for a while. But would he ever take a company public again?

"No, at least not with me as CEO. I don't think I'm particularly good at it," Mason says. "Even in the best conditions, you spend a significant amount of time doing things other than making stuff."

And making stuff--slowly and methodically--is exactly what he wants to be doing right now.