New Total for Online Holiday Shopping: $8 Billion

Jan 15 1999

Could we have spent $8.2 billion on Web Christmas?

That's the 1998 online holiday shopping total, according to a new Marketing Corporation of America study released Thursday.

"We think [the figure] is pretty significant, because a year ago it was only a fraction of that," says Craig Johnson, MCA partner and head of its e-retail practice, who estimates only $1.2 billion was spent online during the 1997 holiday season.

The $8 billion figure is significantly higher than other estimates released to date. Prior to the holiday season, analyst groups forecast sales anywhere from $2.3 billion to $3.5 billion ).

But as the season progressed, many observers - e-commerce stock investors included - felt 1998 revenues might exceed expectations.

The first results came in just after Christmas: Boston Consulting Group reported digital retailers rang up 230 percent more holiday revenue than last year, with average orders up 6 percent to $55.

Then New Year brought hard numbers from the big guys: AOL members spent $1.2 billion on their shopping services, Amazon.com reported revenue of $250 million in the fourth quarter, Barnesandnoble.com holiday sales brought $17.8 million in, and CNET said the site generated $80 million in fourth-quarter revenue for its partner merchants.

But could all this really add up to $8 billion?

MCA says it does. Johnson attributes the low numbers from other firms to methodological differences.

"Other folks are not going to the consumers," says Johnson. "They are making estimates based on a variety of e-retailers' [data] and most don't divulge what they're getting on the Net."

It's true that some analysts, such as Boston Consulting Group, base their forecasts entirely on supply-side research. This traditional means of industry analysis involves surveying the digital retailers themselves to learn how much they plan to sell or have sold over a certain period.

But Forrester Research and Jupiter Communications based their holiday forecasts on both - they used consumer research as well as retailer surveys.

MCA, New York researcher Cyber Dialogue and many other firms base sales forecasts solely on demand-side research. This method involves surveying a scientific sample of consumers to learn what they want to buy and/or how much they bought for the holidays.

Both methods have weaknesses. Supply-side research tends to underestimate totals because it can be nearly impossible to get data from every online merchant, and certain merchants just won't tell. Consumer studies can overstate things, because survey respondents often inflate their activity when they respond to questionnaires.

The MCA study is based on more than 2,000 interviews. Johnson says the firm applied factors to adjust for overstatement, and used sampling techniques to ensure that the findings can be projected to the U.S. adult population, with an error of plus or minus 3 percent.

But MCA also reports 102 million adults have online access - a much higher figure than other researchers use. Other firms prefer to focus on those who use the Web on a regular basis. For example, Media Metrix pegs the number of Americans who used the Web in the past 30 days at 56.8 million, while a Nielsen/CommerceNet study has the next-highest estimates 70.5 million.

Though conducted through similar methods, Cyber Dialogue's post-Christmas analysis estimates approximately $3.4 billion changed hands online. Cyber Dialogue analyst Scott Reents says his firm's holiday sales total could reach "somewhat under $8 billion" only if results included product sales that were shopped for online and then purchased by phone or in a physical store.

But MCA's Johnson insists the $8.2 billion figure includes only sales transacted online.

And while the MCA study reports 25 million adults shopped online during December alone, Cyber Dialogue estimates only 18 million shopped online during the entire year of 1998.

"That just seems way high compared to what we're finding," says Reents.

Correction: Forrester Research and Jupiter Communications based their holiday e-commerce forecasts on both retailer and consumer surveys. An earlier version of this article misstated their methodologies.