By DAVE GUSSOW
© St. Petersburg Times, published December 14, 1998
he bots are busy this holiday shopping season.
An automated army of software helpers is marching around the Internet, helping online shoppers find and compare products and prices, and giving merchants a hand in customer service.
Bot, short for robot, is a computer program that operates as a user's agent. A shop bot is an automated comparison shopper that supposedly trolls for the best bargains on the Internet.
But it pays to know your bots. Some search the entire Internet; some just selected sites; some just information that companies pay to provide. In other words, the bargain you think you've found may just be somebody's covert advertising.
"We're in the Wild West of intelligent agents when it comes to electronic commerce," said Marcus Zillman, a bot expert and chief executive and founder of BotSpot (www.botspot.com). "All we're going to see is improvement."
Zillman's site tells all you ever wanted to know about bots and intelligent agents, including lists by categories, such as ones that will talk to you. Bots and intelligent agents take an Internet search to a higher and more specific level than traditional search engines. They are automated, can filter the information to very specific terms (thus eliminating searches that can generate thousands of hits) and report back to the user.
Bots are not the only thing booming in online retailing this holiday shopping season. Retailers and analysts expect a record year for e-commerce, while conceding that it remains just a tiny portion of the overall U.S. retailing market.
Ken Cassar, an analyst with the market research firm Jupiter Communications, gives three reasons for the surge in online commerce:
* An estimated 75-million people are online this year, compared with 59-million a year ago.
* About 22 percent of those online are shopping, up from 17 percent a year ago.
* Those who are shopping are spending more, about $620 annually, up from $350 a year ago.
Overall, Jupiter estimates that e-commerce will total about $7-billion this year, up from $3-billion a year ago. By 2002, Jupiter projects online commerce to increase to $41-billion.
For the holiday season alone, Jupiter expects sales of $2.3-billion, up from $1.1-billion a year ago. To put that in perspective, if this year's holiday projection is met, all the shopping online will about match the $2.24-billion that retailer TJMaxx did during last year's fourth quarter.
Some categories are beginning to build a following online, Cassar said: 7.1 percent of the sales in the total personal-computer hardware market is online, led by Dell Computer's success; 2.4 percent of the book market is online, led by Amazon.com; and apparel sales are expected to boom, reaching $330-million this year, up from $103-million last year.
Another element driving this year's sales is more brand-name companies online, such as macys.com, and ad campaigns that have driven sales for sites such as eToys.com.
"The brands whose names we are now familiar with are probably the ones who are going to be making the sales this year," Cassar said.
Brand names give a comfort level for online shoppers, said Lauren Freedman, president of the e-tailing group, an e-commerce consulting firm.
"The fear factor isn't the same as from a virtual retailer" the consumer may not be familiar with, Freedman said.
Consumers shopping online are not doing so because of prices, Freedman said. "They're buying convenience. I don't think prices are better or worse. I think they're finding an alternative way to shop. They've got the list, and know where to go for it."
Retailers see a change in consumer attitudes toward online shopping.
"The consumers are really accepting the online media as part of everyday culture," said Stuart Spiegal, vice president and general manager of iQVC, the online arm of the TV retailer. "It's becoming a much more natural act to go online and shop and find information and add value to their life."
iQVC (www.iqvc.com), which was profitable last year, expects sales almost to triple this year, Spiegal said. Part of its success comes from continuing to tweak its site, adding features and services, such as a jewelry site, a feature finder, a gift site and a customer service site, Spiegal said.
"We're just trying to build a better mousetrap," Spiegal said, "concentrating on making our store the best place to shop in."
America Online, the biggest online service with 14-million subscribers, has built an online shopping channel with 110 retailers as partners, spokeswoman Wendy Goldberg said. AOL doesn't disclose sales figures, but Goldberg said AOL is comfortable with projections that overall online holiday sales will double this year.
AOL's goal is to provide a one-stop shopping service, emphasizing convenience.
"People don't want to search," Goldberg said. "They want to find."
But the Web is a crowded place, with thousands of places to shop. Consumers need help in sorting it out, which is where the bots come in, flawed as they may be.
Shoppers' online behavior is well-suited for bots, according to Cassar of Jupiter Communications. In a May survey, 77 percent of the computer users questioned went online to make a purchase. Of that group, 80 percent visited more than one Web site before buying.
"The bots have not done a great job so far of categorizing things neatly, of producing output that's very useful to a user," Cassar said. "They have a way to go before that happens."
BotSpot's Zillman is optimistic that a better bot is not far off. For one, he expects the industry to adopt a single programing language so any bot can understand any data base. But hurdles remain.
"We're beginning to see more and more people put more information and stores online," said Zillman, who is based in Marco Island. "But some merchants don't like bots because bots do compare prices. If they don't have the low price, they won't get the business."
Zillman expects the technology to evolve so that individuals will have personal bots that will be programed to find everything from merchandise to information for the user, bots that will learn the user's habits by remembering frequently visited sites, for example, and bring that data back automatically.
But that is down the road. For now, it's the holidays, and shopping is getting a lot of attention. For consumers who may not know the limitations of a particular bot, Zillman has a suggestion:
"By using two bots, you'll quickly find out that one bot will bring out only (a merchant) who paid for the preference, and another will bring back everything," Zillman said.
"The bottom line is comfort level and the price and quality level of what it brings back. If the content is what you want, then the hit has performed its function."