Flite Is Out to Build a Better Widget For the Internet Ad Industry
By early 2009, Widgetbox was a company on life support. The startup ran a platform that allowed developers to build and distribute small software applications called widgets that could be dropped into consumer or enterprise websites. But the three-year-old San Francisco firm was nearly done in by the triple whammy of the recession (which meant there were few paying customers for widgets), Facebook (which had created its own massive market for embedded Web applications, sucking away software development talent), and Apple (which did the same on mobile devices). The company laid off half its staff, and spent the better part of a year searching for a new business model. “To be totally transparent, we had no idea what to do,” says CEO Will Price.
Then, on July 21, 2009—coincidentally, the 40th anniversary of Neil Armstrong’s walk on the moon—“I got an e-mail that changed the company,” Price says. The note was from Ryan Roslansky, the director of product management at LinkedIn. Roslansky explained that LinkedIn was planning to start showing display ads alongside its social-networking profiles. But it didn’t want to show standard banner ads—it wanted to offer advertisers a different type of ad unit, something that would be more customized, interactive, and engaging, and would therefore fetch higher rates. Such ads would need to function like mini-websites, but they’d have to be far faster, easier, and cheaper to create than traditional “rich media” ads. Roslansky’s suggestion: Widgetbox already had the technology to do this.
“Ryan said, ‘If you don’t call these things widgets but call them banner ads, you guys are on the cutting edge of what ads are,'” says Price. “So we found ourselves in the display ad market.”
It was an epic pivot, and it saved the company. Today the startup is called Flite (it abandoned the Widgetbox name in March), and it’s got a long list of publishers, ad agencies, and brands, including LinkedIn, paying for what it calls its “cloud-based” advertising platform. The “cloud” part is really just the fashionable term these days for what Flite has been doing all along—that is, connecting a gallery of small applications with a back end of data and services. If you see a Web ad for HP, FedEx, or Toyota that has lots of clickable tabs, videos, forms, sliding images, and social sharing links, chances are the ad was authored on, and served up by, Flite’s platform.
Widgetbox ran through $14 million in Series A and B funding before its stall. But with 65 employees and $12 million in new cash from General Catalyst, Sequoia Capital, and Hummer Winblad, Flite is airborne again. Its main challenge now, Price says, is the “behavior change issue.” Big brands that want nifty interactive ads are accustomed to paying big money to big advertising agencies, who put big creative teams on the case and take six weeks to build one ad. None of that is necessary with Flite’s ad-authoring platform, which makes assembling and staging a new ad as easy as clicking a few buttons. Indeed, the system is designed to be so easy to use that brands themselves can change the content of their ads every day, if they feel like it. “Any technology change that requires a behavior change has a risk,” as Price puts it. “But somebody is going to win this ability for brands to do their own programming.”
Oh, and Ryan Roslansky? His 2009 note and Price’s response was the purest possible example of a startup listening to a customer, even if it was driven by desperation. “I wanted to give him stock, but his general counsel wouldn’t let me,” says Price. Instead, Roslansky serves today on Flite’s “Customer Council,” a group of advisors from top publishers, agencies, and brands.
Flite’s story starts even before widgets. Co-founder Ed Anuff had previously started Epicentric, a dot-com-era maker of enterprise portal software, which grew to more than 300 people and attempted to go public before being acquired by Vignette in 2002. Epicentric and its competitor Plumtree, eventually purchased by Microsoft, were both built around the idea that a single Web page—say, the supply chain management page in a corporate intranet—might include several software modules, each talking to its own back end. Widgetbox, which Anuff started with Giles Goodwin and Dean Moses after leaving Vignette, took that idea from Epicentric and turned it toward consumers. The company wanted to build a catalog of widgets—say, a Flickr slide show widget, an eBay auction widget, or a New York Times headline widget—and let Web publishers find, configure, and serve up specific widgets for their audiences.
At the time, Will Price was a Hummer Winblad partner. He says he “fell in love with the concept” after working on Widgetbox’s Series A and B rounds. He was brought in as CEO in 2008, after Widgetbox had built a huge gallery of widgets (half a million—which is more widgets than there are apps in the iTunes App Store) but before it had figured out how to monetize them on a large scale. “The vision was eventually to build a marketplace for Web widgets where someone would say ‘I need a pollen count widget that tells me the pollen count in my school district’ and put that on the school’s home page, and pay a dollar a month to the developer, that kind of thing,” says Price. “We never put in a transactional layer where people would pay for them, but the technology was there.”
But by 2008, it was too late anyway. “The concept of being able to build a viable app business that attracted really high-caliber developers was supplanted by two massively scalable markets that emerged—the ability to build iOS and Facebook apps,” says Price. “They had the devices and the users, a lot more than we had. We realized that Plan A was not going to be a venture-return company.”
Unfortunately, there was no Plan B—at least, not until Roslansky suggested that the widget technology could be used to reinvent display advertising on the Web. “It sounds incredibly stupid to say it now, but it had never dawned on us that an industry that had been around for 17 years and had billions of revenue was ready for disruption,” Price says.
There were at least two big problems in the display-ad business that needed fixing. First, CPMs (the cost per thousand impressions) for standard, static banner ads were dropping precipitously, in part because the ads were increasingly ineffective. The number of people who actually click on them, and then buy products from the advertiser, is abominably low. That was driving publishers, advertisers, and agencies to explore new, more lucrative types of ads that contained “rich media” such as audio, video, or even 3D interactive models, or that included Web elements such as newsletter signup forms or click-to-call buttons for contacting a salesperson.
But that led to the second problem: building traditional rich media ads is slow and expensive, with turnaround time measured in weeks and costs measured in the tens of thousands of dollars per campaign. On top of that, most rich-media ads depend on the Flash multimedia format, which doesn’t work on iPhones and iPads, and it’s difficult or impossible to update their content without starting over.
Widgetbox’s technology offered a way out. “The synapse that connected was that we had widgets for slide shows, click-to-chat, and all those components, and if we just hooked up a series of components as a display ad, we could transform industry thinking,” says Price. “Instead of thinking of a display ad on Site A as a portal to a landing page on Site B, you can think about an ad as a mini-website where people can directly engage with a brand.”
Widgetbox already had an extensive infrastructure for serving up configurable software elements at large scale and monitoring usage. All it had to add was an authoring interface where brands could drop in their own videos and other media and arrange them within the ad unit—see the video below for a taste of how the ad authoring tool works. The process was so much simper than building traditional rich-media campaigns that Widgetbox quickly signed up CBS, MSN, Bloomberg, Forbes, Federated Media, and other publishers as paying customers. As Flite, Widgetbox has fully transformed itself into a “Platform as a Service” company where users pay for access to the tools and analytics; the startup also stores the ads on its cloud servers and serves them up on demand.
In a weird way, Facebook—which had almost killed Widgetbox by drawing away software talent—is now helping to educate the same customers that Price wants Flite to reach. “In traditional Internet advertising, you have a brand, a brief for a seasonal campaign, a creative agency that builds some assets, and a media buy that sends them across the Web,” says Price. “This is a serial process that takes maybe two to six months. But with Facebook, a brand manager can come in in the morning, post a video to the fan page, share a picture, publish a poll question. It’s very easy for them to program content for millions of people. The brand marketing industry has taught people that you can do intra-quarter marketing, but Facebook has taught people that you can do intra-day marketing.”
As an example of this kind of intra-day marketing on Flite’s platform, Price points to a major technology company that offers a free Web-based e-mail system. (Price couldn’t name it, but its identity isn’t hard to guess.) “Today, you market e-mail accounts by making a banner that says ‘Get 5 gigabytes free’ and it takes six weeks, so everything you put into the market is six weeks old. But say Roger Federer lost at Wimbledon this morning—with Flite you can run a banner that says ‘Federer out at Wimbledon, sign up to e-mail this video to your friends.’ Tomorrow, that is going to be old news. The fundamental concept is that once you have a set of Web-based tools that make programming advertising as easy as programming a Facebook fan page, brands can become more agile and iterative.”
That may not be a message that everyone on Madison Avenue wants to hear. Price says the Flite platform drastically reduces the overhead that goes into building or refreshing an ad campaign. Translation: the role of agencies in preparing Internet ads could be drastically diminished, though someone still needs to shoot videos and write copy. “Where brands are going to need help from agencies is being smart about audiences and how much [advertising] do you buy,” says Price. “But I don’t think they have a lot of value in building the current creative that they’re producing.”
If there’s a weakness in Flite’s model, it’s that the finished ads—which are almost always 300 pixels wide and either 250 or 600 pixels high, to fit in the standard slots for Web ad units—have a cookie-cutter feeling, featuring the same formula of videos, tabs, and buttons no matter what brand is being featured (tour Flite’s ad showcase to see what I mean). Even Don Draper might have trouble making them sexy. But it may be that Flite canvas simply hasn’t yet found its Rembrandt; the first agencies that figure out how to put Flite’s platform to more creative use could be in a good position to offer their own clients a more flexible, responsive set of services. McCann Worldwide, one of the world’s largest ad agencies, is already a Flite customer, so “there is evidence that the creative agencies want to embrace this change,” Price says.
Revenue at the startup quadrupled between 2009 and 2010, and is on course to more than double between 2010 and 2011. Price didn’t share the actual numbers, but he says the company is not far from being profitable. And given the scale of the potential market, there’s a lot more upside for Flite’s investors now than there was in early 2009, before that fateful e-mail from LinkedIn. “I think there is a chance to totally redefine a $20 billion market,” says Price. That’s what you call building a better widget.
Here’s a video illustrating the real-time editing capabilities of Flite’s ad platform.