December 11, 2005
Don't Cheat on E-Commerce Search
By Evan Schuman, Ziff Davis Internet
Opinion: As e-commerce sites get more sophisticated about search, the temptation to manipulate the results is overwhelming. Giving in to such temptation is a really bad idea.
Trust and credibility—often associated with a brand—are arguably the most powerful assets of any e-commerce site.
Without it, the Web is just an electronic version of the Wild West, with the shootouts won by the lowest-price site, until that site is shot down by another low-cost site.
Consumers and B2B shoppers purchase from a branded site because they have faith they will be treated fairly and honestly.
Some of the first e-commerce giants to figure this out were some of the larger car company sites, who often placed car-comparison engines on their site. Those comparisons would not infrequently say not-so-nice things about the site's cars.
Why would a Webmaster allow such content on a site? Simple. The customers have easy access to it anyway and they'd rather not give visitors a reason to click elsewhere.
Let's now fast-forward to search. In conversations with lots of B2B IT execs, I hear people complaining about the accuracy or effectiveness of site search engines. A popular request: "If those sites could have a search as effective as Google."
What they don't know is that such a option exists today, and it's free. Go to Google's engine and conduct your search limited to the site you care about. That's the reason so many sites are offering Google search on their own sites because consumers can get it anyway.
But site-specific internal search engines can do so much more. Therein lies the problem.
This week, a search database vendor was describing all of the wonderful things their clients could do for their searches, such as integrating it with inventory so that overstocked items get pushed to the top of search results.
That same philosophy could identify products with lower margins and bury them at the bottom—or exclude them entirely from search results.
It could be married with cookie identification so that customers who bought one obviously overpriced product could be pitched on other items appropriate for those with too much money on their hands.
Such capabilities are mouth-wateringly attractive to some marketers. The entire purpose of this column is to beg them to resist temptation. Monkeying with search results is as bad an idea as if a major car vendor tried to remove anything not supportive from a car comparison service.
In short, you don't want to give site visitors a reason to go offsite and use Google or to leave Toyota.com and go to use Edmunds.com. They have the ability to do that and to do it for free, so don't tempt them.
The counter to this argument is, "What's the harm? I'm not having our systems doing anything that our sales reps on the floor of our brick-and-mortar aren't doing."
That's true, but those reps are not likely to find that technique as effective as it was a decade ago. Consumers today do more research online and won't as easily fall for sales pressure on an overstocked item. If that consumer is looking at merchandise and keying things into their Web-enabled PDA, that sales rep is in a heap of trouble.
The Web's free flow of information gives merchants a reason to be honest and to sell less and explain and help more. By doing so, they'll likely sell more.
Keeping store reps—and site search engines—honest, comprehensive and helpful is the best you can hope for. Cheat them on that and it whittles away the only advantage you have: credibility from your brand.
Interestingly enough, Consumer Reports just did a shootout with some of the largest shopping bots and found the shopbots associated with the two kings of search (Yahoo Shopping and Google's Froogle) delivered the lowest prices, but both fared poorly in safeguarding consumer privacy.
Clearly, doing well in both areas is the ideal, but if they have to make a choice, I think Google and Yahoo made the right call.
Will they be alone?
|