The twice-yearly new product orgy called Demo 08 kicks off tomorrow, but we have a few previews of presenting companies we wanted to write up before the CEOs take the stage. First up: SkyFire, a browser for smartphones.
CEO Nitin Bhandari told me that his goal with Skyfire is to do "true desktop rendering," including media support, on a tiny screen.
Bhandari showed me a demo during our Skype call, and it appears that his browser does just what he intended. Displaying a sports site, Skyfire played the video and audio on it very well. The browser also recognizes text and reflows it into a phone-size column when you zoom into it, so you don't have to scroll side-to-side to read.
The technical trick of Skyfire is that it's a proxy browser. The Skyfire app itself isn't a full browser. Instead, big Skyfire servers elsewhere process Web pages, including all the media and browser formats--like Flash, Java, and Ajax--that a normal desktop browser would handle but that most phones can't. Then it streams data to phones, which the mobile half of Skyfire displays. To the end user, it looks like a browser, but the mobile app is just one part of the product.
This means that Skyfire can do things on mobile phones that mobile browsers, including Safari and Opera cannot, at least not without bogging down the phone. The danger is that the Skyfire servers themselves will bog down, because of over-use, and ruin the experience for mobile users. This is the experience I had with Micrsoft's Deepfish, a proxy browser project we covered last year and that appears to have gone dormant. Bhandari told me the service is "built to scale," and that "once at escape velocity," the company can plug in additional servers to handle a growing user base. He would not reveal the technical underpinnings of Skyfire beyond that.
The escape velocity that Bhandari refers to includes not just user uptake, but a revenue model, and it's on that second small detail that my confidence in this product begins to wane. Bhandari may "monetize user activity," which means selling ads, and he also hopes to generate revenues from carriers that want a competitor to Apple's iPhone browsing experience. The first revenue model is flawed--in-browser ads on a tiny screen will be annoying and hard to sell. The second model is sound, but incredibly difficult. Many mobile app companies have withered and died while waiting for a good carrier deal. And in this case, the carriers are going to need extra convincing, because supporting Skyfire means running or paying for a bank of proxy servers.
Although I think that proxied browsing is the right solution for mobile devices, I am not convinced that there's a solid business behind it.
Skyfire will go into public beta in a few weeks for Windows smartphones. Other platforms will follow. You should be able to register for the beta this week at the site.