The Disintermediation Blues On the sad state of online car- and mortgage-buying services



Hal Plotkin, Special to SF Gate
Thursday, July 18, 2002

URL: sfgate.com/cgi-bin/article.cgi?file=/gate/archive/2002/07/18/disintermed.DTL

Back when the dot-com boom was peaking, there was a lot of talk about how online services would blow away their anachronistic bricks-and-mortar competitors. Some of that has happened. Amazon.com, for example, has played a key role in forcing many local bookstores to close up shop.

But many other targeted lines of business have been far less successful. The missing ingredients in many of those areas are top executives who understand the human, rather than the technical, factors required for success.

Unfortunately, we probably won’t see much progress on that score unless there is a major sea change in Silicon Valley, where technicians such as computer programmers and engineers have long been venerated over workers with the other supposedly “softer,” more people-oriented, skill sets that are obviously needed now.

It’s not news that in recent years people with expertise (or, heaven forbid, degrees) in the social sciences were usually ignored when it came to handing out the top jobs in Silicon Valley, which more typically go to engineers and computer programmers, some of whom also have business-school credentials.

But getting Silicon Valley off its back is going to take a different brand of leader: executives who have a far better understanding of behavioral sciences such as sociology, anthropology, economics, psychology and market research. Sadly, these are exactly the type of workers most Silicon Valley companies hired last, if at all, and fired first.

In fact, Silicon Valley’s tilt toward management by the techno-elite does a lot to explain the failure of the Internet’s promise to disintermediate industries more complicated than the online book and CD market captured successfully by Amazon.com.

You remember “disintermediation.” The word refers to the way new online software-driven business models were supposed to eliminate the middlemen in many industries and provide consumers with better deals they could negotiate right from their computer desktops.

The sorry state of online mortgage and auto-buying services may be the two best examples of the failure of that promise.

Both areas were originally seen as particularly promising, in part because of the super-size transactions involved. Big-ticket items such as mortgages and car sales were supposed to provide the cash flow that would lead to decent profits. Hundreds of millions of dollars were raised from eager tech investors looking to get in on the ground floor of the Internet’s pending restructuring of those industries.

The companies making the software that was supposed to make that happen — sometimes known as configuration engines — were some of the Internet’s hottest properties. They included firms such as Selectica Inc., Calico Commerce Inc. and FirePond Inc., all of which had explosive initial public offerings.

All three of those company’s stocks have since tanked. Selectica’s, which hit $97 a share on its opening day of trading in March 2000, is now under $4, and shares of Calico and FirePond are selling for 26 cents each as of last week.

The reason: Almost none of what was originally promised has happened.

What Went Wrong

The online car-buying services are a particularly vivid example of what went wrong. The original idea, the one that really got investors salivating, was that consumers would be able to go online, visit a car manufacturer’s Web site, pick out the exact model and features they wanted, get prices from dealers in their areas and buy the car from the dealer that offered the best value.

But for the most part, the online car-buying sites that exist today are little more than elaborate ads. When my wife and I recently went looking for a car online, for example, what we most wanted to know was pretty simple: Is the car we want in stock, where is it and what prices would various dealers charge?

What we got instead were the equivalent of online brochures — along with referrals to car dealers in the greater Bay Area. The former we could have easily obtained in the mail, and the latter, from the telephone book.

In essence, because of human factors, little has changed from the first time I bought a new car some 20 years ago. There were the same silly little lies (“The car you want doesn’t come with a sunroof, so you have to buy what we have” or “This is the last one of its type in the state”), the long waits in the salesperson’s dingy little office while the all-powerful sales manager was consulted (“to work out a special deal just for you”) and the prices that went up when we explained that, yes, we actually did expect the car to have tires and a windshield.

Not to mention the fact that, for some reason, whatever car we wanted always seemed to be located in some distant place such as San Diego or Bismarck, North Dakota, which meant additional “destination” charges — in varying amounts — would have to be added to the tab.

The idea that a software programmer would be able to undo all that crap with a few clever lines of code and a neatly designed graphic user interface seems ludicrous in retrospect.

What our experience demonstrates is that no matter how good technology or software may be, what’s really needed to disintermediate an industry is something less technical and far more fundamental: a clear and accurate understanding of the human and economic factors that determine why a particular industry functions the way it does.

In auto sales, for example, although it may not be obvious, dealers add real value to the car-selling food chain, particularly for automakers. For starters, independent dealers relieve auto manufacturers from having to shoulder the full costs of maintaining thousands of retail outlets. They also give consumers a place where they can test-drive cars and kick the tires. The dealers take on these expenses in exchange for a chance to make money selling and servicing vehicles.

Different dealers have different cost structures. Some, in good central locations, pay higher rents than others and must charge higher prices to make a profit. Others sell more cars than smaller competitors, which can win them wholesale discounts they can pass along to consumers. There are many other reasons for sizeable price difference from one dealer to the next on the exact same car: incentive programs, local surpluses of some models, different commission structures for salespeople and others.

The big automakers have absolutely no incentive to sell cars over the Internet in a more open and straightforward manner. And they won’t, not until some clever entrepreneur figures out how to use the technology that already exists in ways that provide manufacturers with something of value similar to the economic buffer they get from their current arrangements with dealers. That’s why it doesn’t matter how nifty some new software is or how many venture capitalists jump up and down about it. What matters is finding strategies to use technology in ways that benefit buyers and sellers equally. It seems unlikely a good Java programmer is going to be able to figure that one out.

A variation of the same thing is true in the underperforming online mortgage arena.

We recently tried one of the most heavily advertised online lending facilitators. You’ve probably seen its ads, which show a would-be borrower kicking bankers out of her office one by one as she shops for the best deal. The lenders, we’re told, will compete for our business.

When we signed up, we quickly got three identical offers, right down to the points we’d need to pay to reduce the interest rate.

But once again, as in the case of auto sales, a human intermediary (our local broker) added value to the transaction, in the form of a reputation for delivering on time as promised, which the online services couldn’t match.

What those designing the business plans for online lenders don’t appear to have appreciated is that when it comes to buying real estate, having a loan approval in hand from a firm with a solid and well-known reputation for delivering is absolutely critical, particularly when would-be buyers are going up against multiple bidders. Sellers understandably want assurance the deal will go through. So, if you were a seller, which offer would you take: one that was backed by a local mortgage broker with a reputation among real estate agents for delivering on time, or a mortgage promised by a dot-com with a post-office box for an address?

Tapping the Opportunity Ahead

The good news is, there is still plenty of room to use the Internet to change and improve the way big purchases such as these are made.

To reach their full potential, online auto dealers, for example, will have to find ways to convince manufacturers to cooperate with them more fully, perhaps by agreeing to buy a certain number of cars the online dealers could then sell directly to consumers. The online services might also offer to share profits with dealers who agree to make their best prices available online, stand behind those offers or let online customers test-drive their vehicles.

Likewise, online mortgage services would be more useful if they offered some form of insurance that protects sellers against delays in delivering loans, and if they found other ways to compete with what a reliable local mortgage broker can offer.

Sooner or later, some clever entrepreneurs will almost certainly figure out ways to make all that (and more) happen. When they do, that success, and the Internet’s next big wave, will be based not on snazzier technology but on a better understanding of the way people do business, what motivates consumers to pick one supplier over another and the human and economic realities of the distribution chain in each particular industry.

Let’s just hope the social scientists and market researchers who tackle those problems get the same stock options the programmers and engineers got.

Hal Plotkin

CC BY 4.0
This work is licensed under a Creative Commons Attribution 4.0 International License.