In some ways, the story of Sift Media is much like that of founder Jud Bowman’s last company Appia.
While innovating at Appia, Bowman and his team discovered a new way of serving up the most relevant ads to specific mobile app users. But Bowman didn’t have time to fully develop the technology at Appia—an intriguing acquisition offer came in from Digital Turbine and it made more sense to focus on the core technology to get the $100 million deal done.
But after the early 2015 exit, Bowman jumped at an opportunity to take that early work and launch another startup. He’d done the same thing eight years prior, spinning the technology that would become Appia out of his first startup Motricity.
Almost everything else about Sift is a different story though.
Bowman raised just over $3 million to spin Sift out of Digital Turbine in January 2016. It was profitable in February 2017 with just four employees.
Appia, meanwhile, raised $30 million and employed 75 people when it hit $33 million in revenue prior to exit. Motricity raised more than $350 million, employed 500 people and reached $110 million in revenue before its IPO in 2010.
Bowman admits that it feels “weird” to hit profitability at this stage of the company. He declines to share revenue figures, only to say that “the amount of revenue we have coming in is equivalent to Appia at 50 people.”
Bowman has stayed quiet about his early success so far. But a new website, launching today, begins to tell the story of a Durham startup quickly becoming a leader in the mobile advertising world.
In just over a year, advertising campaigns run by Sift have helped brands like Facebook, McDonalds, Pandora, Hulu, Uber, Zillow and Audible earn more than a million app installs.
After a year working exclusively with Digital Turbine’s customers to deliver mobile advertising impressions and app installs, Sift has forged relationships with advertising agencies representing top consumer brands, as well as the seven mobile advertising exchanges now used by those agencies to buy and serve ads across Facebook, Google and other mobile apps and sites.
And with a new employee in China, Sift is ready to grow its platform with the major players there.
At Appia, most customer relationships required human interaction. Appia employed a team of business development people who partnered with app companies that wanted to earn revenue from mobile advertising. Salespeople then made deals with advertisers to serve ads on the apps, which would lead to more of their own app downloads. Bowman describes Appia as working in the middle of that relationship, connecting the two parties with a simple platform.
Though Sift also sits in the middle of the advertiser-ad server relationship, it is all about automation.
Starting with 750 million consumer profiles
Sift, so named because of its ability to “isolate that which is most important or useful” (as the dictionary definition describes), begins with data.
The idea came after considering the implications of a growing set of data around mobile app use. A report by Nielsen in 2012 showed that the average person has 41 apps on his or her phone (Sift’s data shows that number to be more like 50 now). Apps, it turns out, create “a unique identifier of who you are,” Bowman says.
Bowman’s team thought if they could figure out what apps are downloaded to a person’s phone, they could create very personalized consumer profiles that advertisers could use to target messaging.
Part of the secret of Sift is in the way it collects data—from each mobile app user, it aggregates 60 data points (anonymously) which has helped it build a database of more than 750 million consumer profiles.
But the technology doesn’t stop there. Using automation and machine learning algorithms, Sift also has to determine in real time which ads best suit those profiles, bid on them in the exchanges and serve them to mobile app users who fit the profiles.
All that analysis happens in the time it takes for a person to open and start using an app. Sift’s engine is so powerful that it sifts through two to three billion ads a day to land the best ones to meet customers’ app install goals—two billion ads so far have been served and delivered on behalf of Sift customers.
And it can even work without people involved. Over the two weeks around Christmas this year, Bowman shut the company down and let the algorithms do the work. Agencies sent new campaigns via the Sift API and then launched without any human intervention. The highest day of revenue for the company happened during that time.
“My last two companies were successful, and we had incredible sales, biz deve and marketing people,” Bowman says. “But if we end up being successful at this company, it will be based entirely on engineering prowess.”
Timing, team, investors and luck
A few things are playing in Bowman’s favor.
First, some key moves in the mobile advertising industry have validated Sift’s work. In 2013, Twitter acquired a big exchange called MoPub for $350 million—that signaled that mobile exchanges would be the predominant way that brands would buy and serve ads moving forward.
Then, last September, a bootstrapped mobile marketing company called AppLovin was acquired for $1.4 billion by a Chinese investor.
“That gave us motivation,” Bowman says. “This space is bigger than I thought when I started this company.”
Beyond that, the early path to revenue has meant Bowman could focus on building the company rather than fundraising. Sift has Digital Turbine as an investor, as well as several investors from Bowman’s past two companies—Wakefield Group of Charlotte, Piedmont Capital of Greensboro and Aileron Ventures of Charleston. Durham’s Idea Fund Partners is in a Bowman company for the first time too.
According to Idea Fund’s Lister Delgado, who admits his investment was mostly due to Bowman’s track record, Sift has surpassed his early expectations for the company.
“It was a big, big bet. Nothing would tell you this was going to work,” Delgado says. “But that feeling has definitely changed. He really is onto something, and something big.”
According to Delgado, “it might be one of the biggest revenue-per-employee companies that we invest in.”
Bowman says he will add to the team (he added two employees in March), but it won’t be comprised of salespeople and it certainly won’t be as large as in the past. He explains his future staff as “rockstar computer scientists and data scientists”.
For now, he’s got his former rockstars from Appia proving they can serve major customers and make money from accurate data and technology.
He’s also got a continued streak of luck in his favor.
“I’m motivated to see if I can be lucky three times in a row,” Bowman says.