Click, Click, Click
Holiday e-retail sales rack up another stellar year, in spite of some performance problems
By Kurt Peters
It was a milestone in online shopping—e-retail sales for the week that ended
Dec. 7 exceeded $2 billion for the first time—26% above sales for the corresponding
week a year ago, according to comScore Networks Inc. numbers. And it was also
symbolic of the entire online shopping season.
From the beginning of November until mid December, comScore reports that consumers
spent $9 billion in online retail shopping, 28% higher than a year ago. Less
than a week before Christmas, comScore was projecting total November and December
online sales of $12.1 billion to $12.6 billion, a 25-30% increase over a year
ago. Meanwhile, total retail sales for the year appeared by mid December to
be on track for 5-6% growth. “With the Internet, consumers now have the capability
to shop with surgical precision and it gives them a level of confidence,” says
Dan Hess, vice president of comScore. “They’re procrastinating this year to
an extent that would have given them white knuckles in the past.”
But things could have been better, say companies that monitor retail web sites.
Starting the day after Thanksgiving, heavy traffic killed transactions at many
retail web sites, reported Keynote Systems Inc., provider of web site performance
and management services. During the peak shopping period of 11 a.m. to 2 p.m.
Eastern time that Friday, consumers could complete a transaction only 80% of
the time, says Keynote’s E-Commerce Transaction Performance Index. Throughout
the entire day—from 7 a.m. to 10 p.m. Eastern time—the success rate was 89.1%.
By comparison, the success rate on the day before Thanksgiving was 98.4%.
Like at the mall
And the day after Thanksgiving was not an aberration, Keynote says, reporting
that even through the week of Dec. 7 there were times when the success rate
reached only 80%. The overall success rate for that week was at a six-week low,
Keynote says, coming in at 93.68%. It improved to 95.98% for the week of Dec.
14. Keynote defines a transaction as a web site’s ability to allow a consumer
to click through a number of pages and successfully make a purchase. The average
success rate for the last three weeks in November and first week in December
was 94.49% to 96.68%. Keynote measures traffic to 13 retail sites.
Keynote was not alone in finding problems with retail web sites. Gomez Inc.,
which monitors 20 retail sites, found a slowdown in ability to complete a transaction
as the shopping season progressed. By the middle of December, consumers on dial-up
connections were taking 95 seconds to access a site, complete a search, click
on an item and add it to a shopping cart. That was up 5 seconds from early November.
The increasing delay was due almost entirely to increased traffic, says John
Lovett, senior performance analyst for Gomez. “It’s like going to the mall on
a Saturday afternoon; you’re going to find long lines,” he says.
Lovett says the problems arise primarily as customers engage retailers’ back-end
systems. “Home pages and searches are processed fast,” he says. “Retail sites
get tripped up when the customer puts things in a shopping cart and engages
the site in more intricate tasks, like inventory checking.”
Gomez has contracts with 10,000 computer users worldwide to allow Gomez to
test web sites from the computers whenever they are online, through dial-up
as well as low broadband such as cable and high broadband such as T1 lines.
It measures the amount of time it takes to access the home page, enter a term
into the search box, get the results, click on the first result and place it
into a shopping cart before abandoning the transaction. Low broadband users
average 27 seconds and high broadband users, 13 seconds.
More orders, lower tickets
In addition, Jupiter Research studied 239 home pages in December and found
that 14% contained bad links, including delivering the infamous 404 Page Not
Found message at 24 sites. While it didn’t report separately on retail sites,
Jupiter says retail was a significant component of the study.
In spite of such problems, consumers are buying online, as evidenced not only
by the comScore numbers but by other researchers as well as online merchants.
Shopping and comparison site BizRate.com, for instance, reports that consumers
placed 64% more orders online this year over last, although the average ticket
was down 25% from $136 to $103, says Chuck Davis, CEO. He attributes the lower
average ticket to the lessening influence of computer hardware purchases.
Among
individual retailers who revealed their online holiday shopping results before
Christmas, footwear retailer Shoebuy.com said sales were up 100% in December
over a year ago; outdoor gear retailer Altrec.com reported sales growth of 94%;
home furnishing retailer Bombay Co. reported that web sales were up 60% year-over-year;
and outdoor gear retailer REI.com said sales were growing at double-digit rates,
without being more specific. Internet traffic monitoring and analysis company
Hitwise reported that retail traffic on the Friday after Thanksgiving accounted
for 8.96% of all web traffic, up from 6.74% a week earlier.
Bluefly.com also noted that sales did not taper off by Dec. 17, as they did
last year. “By this point last year, our holiday season was beginning to wane,”
says president Ken Seiff. “But this year it continues to be strong. Traffic
is up, average order size is up and margins are strong as we’ve been less promotional
than ever before.” In fact, the eSpending report from Goldman, Sachs & Co.,
Harris Interactive and Nielsen/-NetRatings reported that by Dec. 15, only 31%
of online shoppers had completed their shopping and 24% of those who intended
to shop online hadn’t even started by that date.
While customers, like Bluefly’s, were shopping later in the holiday season,
they also started earlier. The eSpending report said that 51% of respondents
started their Christmas shopping in November, up from 43% who started in November
last year.
The broadband effect
For the past few years, researchers have also noted that a big spike occurs
on the Monday after Thanksgiving. While that’s still the case, it’s not as pronounced
as in previous year, says Davis of BizRate. He attributes that to the growing
presence of broadband Internet access in homes. In addition, BizRate’s own traffic
has undergone a change as the result of broadband in the home. Until this year,
BizRate experienced its heaviest traffic around 3 p.m. Eastern time. That was
still the case this year, but BizRate experienced a second spike at 9 p.m. Eastern
time. “That’s the broadband effect,” Davis says. “With more broadband in home,
more people are going online from home rather than waiting until they get to
work to use the broadband connections there. On many days, the 9 p.m. spike
was as big as the 3 p.m. spike.”
While retailers can infer from the growing sales and traffic that consumers
like shopping online, another part of the eSpending Report confirms it. The
problems that researchers identified had little impact on consumers’ satisfaction.
42% of respondents said they were “very satisfied” with the 2003 holiday shopping
season, up 10 points from the year before. “Problems will always occur, just
like in a retail store. Given the volume of shopping, the problems were a relatively
low occurrence,” says Lori Iventosch-James, director of e-commerce research
for Harris Interactive. “Retailers deserve a lot of credit for making things
easier.”
kurt@verticalwebmedia.com
The times they are a-changing
It stands to reason that retail channel shift is taking place as online sales
grow at a faster pace than total retail sales. Is online drawing from catalogs,
as some researchers contend (see, p. 30)? Or is it coming from stores? This
year’s eSpending report from Goldman, Sachs & Co., Harris Interactive and
Nielsen/NetRatings says stores.
The eSpending report surveys consumers weekly through the holiday shopping
season. In the first five weeks of December, 6,000 respondents reported the
proportion of gift dollars they were spending online, in stores and through
catalogs. The results:
2003
2002
Online 20.5% 16.8%
Stores 72.7% 75.8%
Catalogs 6.8% 7.4%
“It’s a significant change,” says Lori Iventosch-James, director of e-commerce
research for Harris Interactive. “It’s almost directly coming from stores. That
indicates the level of comfort that people have shopping online.”
