Surrounded by Search Engines: A New Kind of Mobile Search

Over the past decade, search has evolved based on a combination of better algorithms and increasing user sophistication. Local search, for example, has improved with users’ knowledge to use geographic search terms, and with Google’s ability to better index and serve local business information.

While these factors have evolved, one thing that hasn’t changed much is the input. In other words, through improvements in search technology, we’ve kept the same mice, keyboards, and screens (OK, maybe better screens).

The mobile world will be a different story, mostly due to interfaces that are much more limited. Smaller screens and keypads will force mobile product developers to get clever with how we communicate with the device — including how we search.

This has recently involved accelerometers, cameras, and touch screens which form the basis for search interfaces that don’t involve typing (or tapping). These include UrbanSpoon, bar code scanners such as ShopSavvy, and music discovery engines such as Shazam.

Mobile Search Raises Its Voice

Mobile products from Vlingo, Tellme, and Google have similarly taken advantage of the phone’s internal microphone to build voice search interfaces that find and call local businesses.

Voice search could have a bright future in mobile, and even evolve beyond the phone itself. For example, an area known as speech controlled Internet devices (SCIDs) transforms the myriad electronics that surround into their own little search engines.

Sensory Inc. is working on this proposition. It’s spent the past 15 years embedded speech recognition chips in about 60 million products — everything from in-car voice command systems to Bluetooth headsets, to voice controlled toys.

Nothing new here, but we could soon see all of these devices connected to wireless networks. Competition among wireless carriers is quickly driving wireless broadband (4G) rollouts, which will bring us more connected devices — 15 billion by 2015, according to Intel.

This includes everything from coffeemakers to teddy bears — items that don’t fit our typical view of Internet connected devices. In other words, they don’t come with keyboards or screens, making voice a logical input mechanism.

A Matter of Time

Last week I caught up with Sensory CEO Todd Mozer, who walked me through an example: a clock that is completely speech enabled. It can set and tell time, temperature, and a limited set of other variables, prompted by voice command and enabled by its voice recognition chip.

Again, this isn’t new. But when networked to a server-side speech processing engine, the same clock now can tell you the weather in Paris or the closest place to buy an air conditioner. The connectivity also opens the possibility for ad support.

“We can send the query to the Internet, where we can access a much larger database from Web partners,” Mozer said. “Just like when I go to a Web site to look for the weather and see ads, the same thing could happen where ads come back to the speech-controlled device.”

Call to Action

Other possibilities, according to Mozer, include integrating VoIP capability to call people or places from these different devices. This can include ordering a pizza or making a hotel reservation, subsidized by a company like Skype, which can monetize calls in other ways.

“This would be mostly for transactions, so if we want to run an ad and someone wants to connect live, they can just say ‘connect’ and their SCID basically becomes a telephone,” Mozer said. “Giving impressions by voice is OK, but if you can make it actionable, that’s going to be a critical part.”

Sensory has already partnered with Google to connect client-side voice commands (from a Bluetooth, for example) to a Goog411 call that locates local business information. It also works with Microsoft’s Bing 411, which includes a broader index of information like weather and stocks.

“Most major search engines are going after the voice search space because they know it will be big,” Mozer said. “Many don’t yet understand the importance of having a speech recognition client to assist. Because we’re embedded in places like headset, we can bring billions of users into a server every year.”

Bottom line: If Mozer is right about the explosion of SCIDs, you can imagine a day the kitchen, living room, den, and car become search engines in their own right. True, that’s already the promise of the mobile phone, but this broadens the proposition and gets us closer to becoming truly “wireless.”

Join us for Search Engine Strategies San Jose, August 10-14, 2009, at the McEnery Convention Center.

Related reading

Simple Share Buttons