Meet the future: Our cars, ourselves

Chet Huber is responsible for pushing advertisements into the place many people reserve as the last remaining sanctuary of privacy and freedom: the automobile.

So it's not surprising that the mild-mannered, Midwestern engineer heading General Motors' OnStar division is going about his charge in an inoffensive manner. OnStar is conducting focus groups to determine how to beam location-based commercials from banks, gas stations, movie theaters and retailers--but in a way that won't annoy or distract drivers.

OnStar's first automated, location-based service will debut later this month. Consumers will be able to push a button and receive traffic reports covering a 5-mile radius from their vehicle.

Location-based promotions will affect millions of drivers. OnStar already has more than 1 million subscribers, and it is available in most GM vehicles as well as in cars by Lexus, Acura, Audi and Subaru. The division's goal is to be included eventually in about half of the 15 million vehicles sold per year in the United States.

Huber, who usually works in Detroit, flew to San Francisco last week for a string of business meetings. Between them, he sat down with CNET News.com and talked about location-based ads, mounting driver distractions, and how Americans perceive their vehicles in the 21st century.

Q: OnStar's "Virtual Adviser" service already features four-second sponsorships from Fidelity Investment Services when people check their stock prices or trade stock in their car. What more is OnStar considering, and why?
A: We've learned that the phrase "stock quotes brought to you by Fidelity Investment Services" is something that, when done appropriately, adds to the authenticity of the information itself. It doesn't take away from the experience at all and gives a tremendous opportunity to make an impression from allies and partners. We're getting hits on a daily basis right now. So we've started. We're going to expand that, but very carefully, and gauge consumer reaction.

It's not a big stretch to think that car brands themselves would look at ways to touch a consumer in the vehicle--that'd be a pretty interesting way to do it. We see a logical extension--a message that says, "This transaction capability is brought to you by Cadillac DeVille." It's building the equity of a brand that someone's already in.

In January you predicted location-based ads attached to the car's diagnostic system, so the customer who doesn't have much gasoline learns that the Sunoco station has gas cheaper than anything else nearby. How is that effort proceeding?
We're starting to talk to people broadly about location-based advertising possibilities...The infrastructure, the backbone, the technology is all there. Now it's a question of, "OK, what opportunities are there for extension of that with businesses proximate to where you are?" We're listening to people with ideas. Businesses--movie theaters, gas stations, hotels--are pretty interested in those things.

At some point, you would set up your profile and all of the things you're shopping for--maybe not urgently shopping for but they're on your to-buy list--will get bounced against the database. You'll be driving along and it will say, "Oh, by the way, within three miles of where you are now, that DVD player you said you wanted is on sale at Circuit City."

Wouldn't that be supremely annoying to those few remaining drivers who want to drive in peace? Does mobile commerce mean that a bastion of solitude is just another marketing medium?
We're going to have to do a lot of work with consumers and subscribers. First, we don't even know if there is as much to this in reality as analysts have tried to paint for m-commerce theoretically. The other part is, how much of that starts to look and feel to consumers like it has crossed a line? We're going to move up to any of those lines very carefully. We want them to highly value the service, not create something that's constantly tapping them on the shoulder saying, "Here, we've got this deal for you." We're going to really be careful as we pursue the appropriate sensitivity path.

How do you determine where to draw the lines?
A good series of focus groups. We've got the unique ability to do some pilots and...make this thing be pretty microscopically targeted to services we think would work. Getting national coverage--let's say you wanted to do something with movie theaters--it's almost impossible to roll it out all at once. So we'll conduct some experiments on a very localized place.

Newspaper readership is declining, and advertisers are desperate for new ways to reach people in geographic areas. Would OnStar's advertising rates be geared toward local businesses, or would advertisers have to be deep-pocketed national chains?
The more granular you can make the advertising, the more valuable to the consumer. It's not just one more message pushed in the consumer's face that you feel annoyed by. We have to find that spot, to the extent that it exists, where it's so targeted and meaningful to the consumer that it doesn't feel like advertising--it feels like it's helping the consumer do what they need to do.

But then you've got all of the issues of the administrative side. How would you make a targeted program for a small business work? I'd question the cost-effectiveness of having thousands of individual relationships with small advertisers. We've got a lot of learning to go.

Revenue from location-based wireless services in North America will increase more than 100-fold to $3.9 billion by 2004, according to The Strategis Group. Is the spectacular growth of wireless commerce just hype, or can you bank on it?
I don't know what's "bank-on-able," frankly. We've tried to look at this from a number of different directions and get down to the core of what consumers are really interested in: What's the unmet need here that's going to drive demand? We're not interested in simply what is possible to offer them from a tech standpoint.

We are highly optimistic about and focused at the moment on the vehicle environment itself. Whether it's gas stations or fast food or the travel industry, there are some fairly logical extensions for m-commerce here. I think at the same time there are some transactional businesses that could take advantage of the opportunities to use the time that people have in their vehicle in a way that's respectful of the driving environment.

OnStar is a voice-based system. Other automakers have advocated screen- or even touch-based systems, and Ford is still studying driver distraction. Is there any consensus in the auto industry about which system is best?
I don't see consensus. Certainly the partners we're working with understand, in the most basic sense, the challenge of doing this well. It seems to me that there are still some companies and initiatives that think that visual displays in a vehicle make sense or have a roll in the vehicle as an interface device. I honestly believe, with everything we're seeing, that's not a path with long-term sustainability from a public policy standpoint. Screens in the back seat or passenger's seat give you more options. But the last statistic I saw, something like 80 to 85 percent of the miles driven in this country are driven by vehicles with only one person in them. It's not as if this notion of the whole family in the car with the kids in the back seat is happening where most of the miles are being driven.

Is the idea of location-based advertising and onboard communications predicated on the fact that we all spend so many hours in traffic jams and the situation is going to get even worse?
I just read a study that said in San Francisco the average commuter spends 42 hours a year in traffic delays. In Seattle and L.A., it's more than 50 hours. That's a lot of time...You can talk to traffic engineers and they'll say the same thing: It's going to get worse before it gets better.

That's why we believe the traffic information that we're going to start providing very shortly is going to be very, very useful to our consumers. The ability for people to get a little visibility about what's out there and make alternative routes, we think that's going to be a great service. If we can ensure the integrity of the underlying content, we think that's going to be highly valuable.

The notion is to leverage this network of vehicles with built-in technology, so you can create real-time traffic flow patterns without flying helicopters over the highway. The vehicle itself is saying, "I'm on (California Highway) 101 going 55 miles per hour. And somewhere in the back with a Cray computer, somebody has mapped out the algorithms to determine the traffic flow. It's a self-value-generating system that generates value back to the consumers.

That sounds great, but there's a fundamental problem to the idea of a vehicle network: There is no single device or dashboard port to receive and send information; the industry can't agree on a standard. So there won't be a uniform data system anytime soon, right?
We're trying to minimize the dependence on the hardware. A thin-client model is a lot easier than millions of distributed systems. But at some point you always hit the point where if you had this much more processing power you could do X. Or if you had this much more bandwidth into the vehicle, you could do Y.

We're working hard to create a vehicle that has the basic, fundamental connectivity for the thin-client model, but have it be customizable at the point of vehicle purchase or subsequently during the vehicle's life. It's the conceptual notion of a little server rack in the car with X amount of capability built in. Then if someone's built a locally available high-speed network, we give you a little card, and in your glove compartment there will be a rack and you swipe the card through and you made your own value decision.

Another snag for OnStar: Lots of people use handheld computers, wireless pagers and cell phones for traffic updates. How do you compete against Palm, Research In Motion, Sprint and others, especially if you only offer services in the car--not when people are on buses or in the office?
The way we're looking at the portable issue is that we're optimizing our relationship with the consumer around our initial acquisition: the vehicle. We believe the vehicle is an incredibly rich service environment for the consumer. But we know that as our service suite expands, a number of those offerings are going to be very interesting to consumers outside the vehicle. So you're right. It becomes a competitive issue. If we're not going to provide services, it's an opportunity for someone else.

Obviously, air-bag deployment doesn't help you when you're walking down the street, but there are parts of the suite that are voice portal services we can make available to other services. We see the relationship with our subscribers formed around the vehicle, then a certain subset available through multiple devices, whether Palm VII or WAP (Wireless Application Protocol) phones or literally just a cell phone.

GM has been aggressively acquiring tech companies. Especially considering that you could buy them at bargain prices right now, would cash-rich GM ever buy Palm, Research In Motion, Handspring or another potential OnStar rival?
The last thing we'd do at the moment would be to make a bet on which of these devices would get mass application. We don't want our customers...to give up something they already have and are used to using. The ultimate for us would be to say, "Let's make our service suite...agnostic to the hardware device it's going to be sent to."

If somebody is committed to their Palm VII or BlackBerry or Sprint PCS phone, we will allow the customer to take advantage of our services. We want to make a customer's life simple. We're not going to ask everyone to learn one more special platform; that's not going to happen.

The automobile has been a central force in Americans' lives, shaping where we live and even what we eat. How will m-commerce change that?
When I first came to OnStar in 1995, the prevailing world order explained to me by deep thinkers was, "You've got home, office and mobile." It seemed to us at that time that maybe we were overly fixated on the vehicle because we had GM as a parent, but we thought the vehicle was a more special place than just one element of a mobile environment.

People relate pretty heavily to their vehicle. It's a predictable relationship. You're there during certain times, doing certain things--not unlike a home or office. Honestly, I believe that the vehicle, as it becomes more connected, is going to feel like one of those core places in our lives: home, office, vehicle.

You're asking if it's significant that we are creating new services and options for the 500 million hours a week people in the United States spend in our cars. Yeah, I think so. The physical characteristics of the vehicle are absolutely meant for m-commerce. Essentially, it's a power generator on wheels. You never have to think, "I hope my battery doesn't run out. I wonder how much standby power I've got." We've got incredible packaging volume to work with, compared to a shirt-pocket scenario. We've got a very, very special environment. People are not going to think about their vehicles in the same way.