Talk to Me
By Michael Mattis, Tue Aug 07 00:00:00 GMT 2001

Pay attention to voice-enabling mobile Internet apps.

It's no secret by now that mobile Internet technologies such as WAP are simply not living up to forecasts regarding either penetration or profits. This, pundits say, is due largely to the usability problems inherent in WAP - slow downloads, faulty connections, frustrating information architectures as a result of small-screen layouts, to name just a few.

That's why a bevy of companies have geared up to deliver net-based apps through the mobile phone's premier interface - voice. Companies like TellMe and BeVocal have been offering voice dialing and other services for some time, allowing telco users to get phone numbers and other information just by talking into their phones.

Now, thanks to advancements in voice recognition and "data scrubbing" techniques, these and other companies are set to voice-enable a whole range of services and applications - heretofore available only online, through the wireline internet and to those willing to lug around heavy laptops with wireless modems.

For consumers, services include calendaring, voice-activated e-mail, restaurant and entertainment reservations, driving directions, and so forth. Enterprises, too, are beginning to rethink the way they deliver sales and field-force automation applications to their mobile work forces. The trick is to be able to voice-enable existing Internet applications without having to develop whole new ones that support the voice platform.

It's a numbers game

Voice-enabling Internet apps is a worthwhile endeavor, say analysts. The Kelsey Group predicts that the U.S. voice portal market alone could reach 45 million users by the year 2005, while voice portal providers could rake in nearly $4 billion. Meanwhile, the market for voice-enabled enterprise apps is expected to top $5 billion in 2004.

Not a bad chunk of change - and let's hope those figures turn out to be less hype-ful and closer to reality than the last century's forecasts were for mobile Internet. (Remember these - sixty-two million wireless handheld users by 2003; mobile commerce users to reach 74 million by 2004; six hundred and eighty-four million microbrowser users by 2006, etc.) The predictions will prove true, say the pundits, only if voice enabling can make mobile Internet applications more usable, not more frustrating.

One Silicon Valley startup, Voice Access Technologies Inc., of Los Gatos, California, thinks it has the problem beat. Founded last year by Peter Olson (Octel, Telocity, and Flycast), Voice Access says it can offer a single solution to both wireless carriers and enterprises that permits users to make information requests over any phone in natural language and get back responses in whatever format is handy - automated voice response, SMS, WAP, Web, even e-mail.

The key, according to Voice Access' VP of Sales, Steve Ghareeb, is the company's "data scrubbing" and preparation engine. This technology, says Ghareeb, enables voice access to prepare and send data to multiple presentation platforms and devices, such as WAP-enabled phones, PDAs, lap- or desktop computers, and as voice responses over the phone, without any intervention from the user. In addition, he notes, Voice Access's system can work within existing telco infrastructures, integrating with signaling system 7 (SS7) billing systems, so there's no added legwork when tracing roaming calls and ensuring that the right customer is billed.

Besides Voice Access, Tellme, and BeVocal, other voice players include, Audiopoint, and ShopTalk. Most of these, with the notable exception of ShopTalk, which specializes in enterprise applications, are targeted at helping consumers get restaurant, traffic, news, sports, stocks, weather, and other "commodity" information.

While these services can be compelling value adds for telcos and portals, voice-enabled applications will likely see the most value within the enterprise, experts say (remember that $5 billion by 2004 prediction), particularly with regard to sales - and field-force automation.

Saving your reputation

Sales techniques, particularly in the U.S., and particularly with big ticket B-to-B durable goods such as routers, servers, fiber-optic cables and cable-pulling equipment, involve a lot of socializing between sales people and their clients.

Generally, the more expensive the purchase, the higher up a sales person must go to get an approval. High-level enterprise players like executives tend to like to know that they are in good company. Sales people know this, and often wine and dine clients as if they're old friends - they chat a lot about family, golf or tennis and so forth before ever getting down to business. It's been done like that for years and no Internet revolution is about to change it.

So where do voice enabled net applications fit in to this old school, cigars-and-martinis picture?

Here's the scenario - say you are a router sales person, selling multiple switches to a major telecommunications infrastructure provider. Individual sales run into the tens of thousands of dollars each. You have built a relationship over the past few years with a C-level executive who makes these purchasing decisions at this major corporation. You like him. He likes you. You're pals who frequent the same country club and often meet on the links for a round of golf and nightcap at the 19th hole.

Your usual modus operandi is to take him out to lunch at a nice restaurant, and then head back to his office to get down to biz. In between chatting about your handicaps, it surfaces that your client's company is expanding, and requires some new equipment. You get the facts, determine his needs, whip out your mobile phone, and get a quote. Your client's perception - that you have someone on the other end of the phone doing the legwork for the sale - is important to both him and to you. Your client is an important person. He needs to feel he's dealing with other important people. It helps him feel good about himself and his relationship with you (and, subsequently, helps the sale and therefore your commission) for him to think you've got what Americans sometimes call "the juice" (i.e., power and authority).

This is where screen-based PDA applications fall down. You could just as easily - and for less overhead than talking to a live person - tick a few boxes off with the stylus on your Palm VII to get the real-time quote. But that would ruin the illusion of having an assistant to do the grunt-work, and make you seem less influential to your high-level client. It may sound silly but it's true.

Now (on the other hand) what if you could speak in natural language to a voice-enabled Internet application that could fulfill your information request without your client ever having to know you were talking to a machine? You could preserve your reputation and save your company money in labor costs at the same time. It's a winning scenario for everyone.

Safety first

Voice-enabling Internet applications has other benefits. Many U.S. states are considering banning mobile device use in automobiles due to alleged safety concerns. Visual-based applications such as those used with PDAs can be particularly hazardous, lawmakers believe, as they force users to look away from the road and take their hands from the wheel. But voice-enabling these applications allows users to employ hands-free sets to fulfill information requests, such as automated voice dialing and calendaring, with little or no distraction. Voice Access's technology, for example, enables voice-activated information requests to be replied to in multiple formats (what the company calls "multi-modal" response).

This enables you to voice-dial for your calendar information, make the request in natural language, and receive your schedule back not only as a voice response, but also as an SMS or as an e-mail to your laptop, which you can then refer to after you've pulled safely off to the side of the road.

Don't count your chickens...

While voice-enabling technologies have come a long way in the last year, it's unlikely in a sagging global economy that telcos and enterprises are going to rush into this new field with the same zeal that they once rushed into the wireline and wireless Internet arenas. But to stay competitive, enterprises with large mobile workforces in particular are going to have to make significant mobile infrastructure investments over the next few years.

The key will be for those enterprises to study carefully what voice-enable mobile services and applications will best serve their field forces' needs. We may not see a voice-enabling free-for-all, but there will be opportunities for applications developers, telcos, and enterprises to create new applications, shave costs by automating business processes and enhance revenues by offering new services down the line.

Michael Mattis is a mobile solutions specialist in the San Francisco offices of Razorfish, Inc.