• by Laurie Sullivan , Staff Writer @lauriesullivan, April 13, 2012

Inflections in a person's voice provide hints of intent -- even more so than a click. But would an engine turn voice into text before matching verbal commands to textual keywords? How much will location-based services and social signals come into play when matching searches with paid-search ads? And will voice search introduce us to radio-like ads, rather than textual -- similar to those on Internet radio Pandora?

Google's voice search has become incredibly accurate on the desktop. Clearly, it relies on location-based services, social signals, and personalized search to serve up nearby Mexican restaurants when voicing the search command "Mexican restaurants."

Brian Kaminski, COO at iProspect, believes the next natural progression will become text, but in the future "if I'm taking advantage of a voice utility, then I probably want audio to come back to me instead of links." Poor return on investment from the ad results when disconnect occurs between the content and the search.

Will the search algorithm detect the inflections in the voice when the person speaks to determine the stage of intent?

Intent, in general, becomes a challenge for search engines, Kaminski said. Another issue becomes harnessing the potential and how quickly marketers will adopt this technology.

Think about the hours people spend in their cars, and the number of automakers integrating Internet access. Car navigation systems already have sophisticated voice systems. Will they use personal assistants like Siri or Evi?

Ford began integrating Internet access for work trucks years ago. Now, Toyota, Nissan, Chevy and other major car manufacturers will continue to introduce connected cars this year. Some will have a search engine, aside from connecting with Facebook, Pandora, and through other apps.

Kaminski said imagine verbally searching for a nearby "restaurant from your car, and using click-to-call to order a pizza." He said there are still technology barriers today causing slow adoption of voice search. Sometimes background noise prevents the microphone from clearly picking up the verbal command.

Google and others will work around that glitch. A Google patent will use background noise from phone calls to target ads.