Monday, December 1, 2008

Black Friday, Cyber Monday and the Importance of Benchmarks

The 2008 retail frenzy is upon us and site operators are busy maintaining online shopping carts while executives are anxiously watching their bottom lines. An early report issued by ComScore this morning indicates that Black Friday 2008 drew in $534 million in online spending. This is a meager 1% bump from last year’s sales. While these numbers are low, it could be worse since myriad factors – including the dismal economy – are looming down on would-be shoppers. But don’t ink those calculations just yet. Today, dubbed “Cyber Monday” by the National Retail Federation is historically a more accurate predictor for online spending. The JupiterResearch US Online Holiday Retail Forecast pins 2008 online retail growth at 12% despite the woeful economic challenges. Still, this is the smallest increase in online shopping since the inception of the Internet, which is poised to blow away off-line holiday sales, forecast at only 2% growth. As an impartial outsider, these figures lead me to ponder the importance of benchmarking within retail and Web site operations at large.

As Americans, we love to compare. Benchmarks within the online world offer us the opportunity to gauge success against industry norms, fierce competitors, and even ourselves. Companies like Gomez, Inc benchmark Web site performance and availability to inform us how well sites stand up against traffic spikes and heavy loads. In typical fashion, this year proved that even the largest brands are susceptible to performance degradation. Today beginning at 10am ET the Victoria Secret shopping cart was producing errors and Williams-Sonoma experienced slowdowns at their shopping cart page – a critical transaction juncture. Over the weekend, Overstock experienced an uncharacteristic slowdown from 5 to 6 second average to over 17 seconds for synthetic transactions. Sears.com suffered the biggest outage, with nearly 40% of transaction attempts squashed on Black Friday. While these numbers are interesting, they are possible through benchmarking. By evaluating performance and availability during non-critical times, the numbers during the holiday crunch can be evaluated in light of what’s normal.

Coremetrics applies this rigor to Web analytics benchmarking. Traditionally Web analytics is an introspective endeavor, with metrics used to compare internal hourly/daily/monthly/seasonally/yearly performance against one’s own measures. Enter the Coremetrics Benchmark Industry Reports that now provide historic comparisons of retail performance across a number of key Web metrics. Data parsed out for the Black Friday surge indicated that 2008 online shoppers; spent less time on sites, converted fewer new visitors and decreased in average order values, when compared to data from Black Friday 2007. A few verticals within retail including department stores and gifts showed promise, but not enough to buoy the entire retail category. While the data may be bleak, the benchmark provides a granular look at metrics that matter for retailers. These data, when evaluated internally, should be used to help site owners determine where to prioritize action on their own sites. An external comparison should be used to illustrate how individual numbers compare against the whole. This tactic may save few merchandisers from the fire, but at least provides context to metrics that matter.

Benchmarking is fairly unique among Web analytics vendors and currently offered by only Coremetrics and Google Analytics. Do you use Web analytics benchmarks? If so, I'd love to hear your use cases and how benchmarks provide valuable insight to your organizations.

No comments: