2011 Holiday Shopping Soars Online
Desktop Web performance is solid; mobile gets mixed reviews
If ever proof was needed of how critical online channels now are to retail success, the results of the extended Thanksgiving holiday shopping weekend provided it. That's plural “channels” because mobile is making a huge leap in prominence this year for holiday shoppers, and tablets—as yet still synonymous with“iPad” —demonstrated the important role they are going to play, too.
And this year more than any before, retailers showed that they recognize how important performance is across all of their online channels. While there were a few notable outages and site slowdowns, absent were the spectacular failures that have been seen in past years. Overall, retailers appeared to be well-prepared, both on the wired Web and to a lesser extent, on mobile, for the shopping frenzy that is as much a part of the Thanksgiving tradition as turkey, football and pumpkin pie.
“Performance was remarkably robust across the sector on the desktop Web,” says Dave Karow, Keynote senior product manager. “The problems we saw were fairly minor and fairly short-lived. We didn’t see anybody fly off the top of the chart in terms of response time, or disappear off the bottom of the chart in availability.”
Three Five Days of Thanksgiving
It used to be that Thanksgiving shopping meant Black Friday and its weekend carryover. Then Cyber Mondayextended the online excitement, last year contributing $1 billion to holidayweekend sales and setting a record for the biggest online retail commerce day in history. This year, both bricks-and-mortar and online retailers stretched the weekend in the other direction, starting off their sales and promotions in a bigger way than ever on Thanksgiving Day itself.
The debut of “couch commerce.”
More than half of online retailers offered Thanksgiving Day-only deals to lureconsumers chomping at the bit to get their holiday shopping started. 2IBM Coremetrics Benchmark, “Black Friday Report 2011” This helped boost online spending on Thanksgiving Day by more than 39 percent over 2010, according to IBM Coremetrics Benchmark. 3comScore press release, “Cyber Monday Spending Hits $1.25 Billion to Rank as Heaviest U.S. Online Spending Day in History,” 11.29.11 All told, online retailers racked up nearly half a billion dollars in sales on Thanksgiving Day—a jump of 18 percent over Thanksgiving 2010. 4IBM Coremetrics Benchmark, “Black Friday Report 2011”
Retail pundits have coined the expression “couch commerce” to describe the phenomenon that includes logging on to do some shopping after the pumpkin pie is finished. Thank the iPad for contributing to this new trend. In fact, on Thanksgiving Day and Black Friday, the iPad was the number two mobile device used for shopping, behind the iPhone and ahead of Android phones. Nearly five percent of mobile shoppers were using an iPad, and their conversion rates were more than 60 percent higher than users of other devices. 5Ibid.
Black Friday booms online.
Many shoppersopted for online shoppingon Black Friday, perhaps to avoid the elbows of shoppers who trained for the Walmartwaffle iron grab. The online channel delivered a retail golden egg again this year, anteing upa nearly 25 percent increase over its 2010 contribution to retail’s biggest day. The big story was the surge in mobile — sales more than tripled to 9.8 percent of total sales volume over 2010, on a 255 percent increase in mobile to 14.3 percent of online retail traffic. 6comScore press release, “Cyber Monday Spending Hits $1.25 Billion to Rank as Heaviest U.S. Online Spending Day in History,” 11.29.11
Some online-only retailers, such as Gilt Groupe, deliberately targeted shoppers who were out fighting the Black Friday lines and crowds.
“Some online retailers are actively promoting and soliciting mobile-only deals during Black Friday,” says Keynote’s Aaron Rudger, “with the intent to basically poach people who are standing in line at physical stores, to try to get them to cross over and pick up deals. You don’t line up unless you’re deal-hungry and deal-driven. They’re standing in line bored, with nothing else to do but play with their smartphones, so some online retailers are taking the opportunity to try to pick them off.”
Cyber Monday sets two online records.
For the second year in a row, Cyber Monday set a new one-day online retail record, handily beating last year’s billion-dollar day by nearly a quarter-billion dollars. More than 10 million people made online purchases on Cyber Monday, another one-day record. 7USA Today, “Buyers click away as Cyber Monday sales jump 18%” by Hadley Malcolm, Jayne O’Donnell and Laura Petrecca, 11/28/11 Overall online traffic was up 43 percent from last year. 8IBM Coremetrics Benchmark, “Cyber Monday Online Spending Increases by 33.0 percent Over 2010”
Mobile was important on Cyber Monday, but at a somewhat lower level than Black Friday, accounting for 10.8 percent of traffic and 6.6 percent of sales. 9comScore, “Cyber Monday Spending Hits $1.25 Billion” Consumers apparently opted for higher-speed workplace connections over their slower mobile links; comScorereports that 50.2 percent of Cyber Monday purchases were made on workplace computers. 10Ibid
Overall, online sales for the five-day period from Thanksgiving through Cyber Monday totaled nearly $3.58 billion, a jump of 20 percent over the same period in 2010. 11National Retail Federation, “Holiday FAQ” 2011 Given that holiday sales are projected to grow just 2.8 percent, 12comScore press release, “$6 Billion in ‘Cyber Week’ U.S. Online Spending Sets New Weekly Record as Three Individual Days Surpass $1 Billion Threshold,” 12/4/11 it’s clear that online is once again contributing more than its share, which is good news for those retailers who are successfully cultivating online shoppers.
Cyber Tuesday, Cyber Wednesday Surprises.
Perhaps the biggest surprise of the holiday kick-off was the addition of Tuesday and Wednesday to the billion-dollar club. Both days joined Cyber Monday in racking up more than a billion dollars in online sales—$1.12 billion Tuesday and $1.03 billion Wednesday. 13Mashable.com, “The Tablet Economy,” by VeenaBissram, 11/28/2011; Verticinfographic with data from various sources Add them to the record book as the third and fourth billion-dollar online days, and as the first three-day stretch of billion-dollar days. If this trend continues, the new official holiday kickoff may end up being a full week long.
Mobile Shopping Goes Mainstream
The holiday weekend brought good news all-around for retailers, but particularly this year for those that planned and executed a solid mobile strategy. Virtually one in ten online buyers — that’s buyers, not shoppers — was using a mobile device on Black Friday. Well-heeled Apple-toting shoppers led the pack; the iPhone and iPad were the number one and two mobile shopping devices on both Black Friday and Cyber Monday.
While many big-name retailers got through the holiday weekend with generally acceptable performance, it’s because they are at the leading edge of an industry that’s largely still trying to figure out its mobile approach, despite a surge in consumer adoption of smartphones. A surprising number of top retailers — including the likes of Macy’s, Costco, Nordstrom, and Bed, Bath & Beyond — posted lackluster performance and availability numbers for the five-day holiday period.
Dell and Sears were the top performers according to Keynote data, with home page load times averaging less than four seconds, rivaling desktop performance. Dell’s speed, however, can be attributed to a lean splash page that directed visitors to the app store to get the Dell app, or to the full mobile site. Not everyone was so fast, though. Of 30 top retailers measured by Keynote, seven averaged home page load times greater than 15 seconds, and five took an excruciating 20+ seconds to load.
“Mobile has become a huge driver in overall retail traffic,” says Kevin Mullen, business development manager for Keynote. “But it’s happened fast enough that many are unprepared for it. Many retailers just barely got out in front of the holiday rush. And not a lot of them were thinking forwardly about mobile load testing. But it should be a standard best practice for retail holiday planning. You load-test your main site. And you want to make sure there are no differences with your mobile site.”
Many retailers are making efforts to enable genuine mobile commerce, as opposed to trying to drive traffic from mobile to other channels to complete the transaction. And many are bringing their mobile development efforts in-house, after testing the waters with a third-party solution. Both of these scenarios involve significant risk, making it all the more imperative to perform robust load testing and ongoing performance monitoring.
“People are experimenting,” Karow says. “There are many who are still just figuring out ‘how well does my site work over mobile? I don’t have a lot of data points, I don’t know how to get good data.’ You don’t know until you know, and it would be better to know before people are trying to get to your site on Black Friday or Cyber Monday.”
The big challenge that mobile sites face that’s different than the desktop Web — in addition to the inherent slowness of over-the-air (OTA) signals — is the plethora of devices, operating systems, and carriers. A mobile site or app has to be vetted in all the major configurations that ultimately control what is displayed in the user’s hands.
“It’s a lot more complex,” says Ben Rushlo, Keynote director for performance consulting. “We’ve found instances where sites are working on one phone correctly but not another, or one carrier much better than another. The added complexity makes testing and measurement more important — and yet we’re not seeing as much discipline in that area as we’d expect for some retailers.”
Speed shopping on the desktop Web
While retailers may still be sorting out their strategies for mobile, on the desktop side they had one of the best-performing holiday seasons yet. Of 47 sites monitored by Keynote during the five-day holiday period, more than half had average home page response times of less than 3 seconds; the overall average for all 47 was 2.96 seconds. All but seven averaged less than four seconds. Traffic peaks slowed sites somewhat, but typically only by one second — a delay that would be barely perceptible to users.
Heavy holiday traffic did cause some availability issues, but no spectacular failures. Ten of the Keynote-measured sites saw some downtime; Cyber Monday saw dips in availability at 7:00 a.m. and 9:00 a.m. and again at 5 p.m. Toys R Us and Dell showed some periods of compromised availability, but not to a catastrophic level. By far, the majority of measured sites were 100 percent available throughout the holiday weekend — an impressive accomplishment given the record-setting traffic.
Any outage or excessive slowness during peak holiday periods can inflict serious pain on a retailer’s bottom line, whether the retailer is big or small, online-only or hybrid. The only way to make sure a website can handle an extraordinary rush of traffic is to thoroughly test it well in advance and every time changes are made. The most solid test regimens subject a site to traffic that’s multiples of the highest demand projected by the sales and marketing department.
“Successful retailers on the Web plan correctly and do their diligence well before the holidays,” says Rushlo. “They’re disciplined nine months out of the year to build a stable platform and a culture that cares about performance and the customer experience.”
Because websites are living things that are constantly changing and evolving, load testing and performance management are ongoing tasks. It may be possible (though not likely) to lock down code during the few critical weeks of the holiday, but the rest of the year sees a string of changes large and small.
“Every year the technology is changing,” Rushlo says. “Whatever gets thrown at you, you have to manage as the site owner and then also have the culture and the discipline to test, measure, certify and make sure that it’s ready. It’s sort of like having to rebuild your entire store every two to three years. A lot of owners are bringing sites back in-house after being outsourced. They’re dealing with new platforms, backend systems. I think most people probably don’t fully understand how much change there is on a retail site.”
The rise of the third screen
Retailers still struggling with their mobile strategy have yet another factor complicating their task: the near-lightening-fast adoption of tablet devices by consumers. As noted above, the Apple iPad was the second most-used mobile shopping device during the holiday weekend, with conversion rates notably higher than other devices.
Tablet ownership is anticipated to double in 2012, to 54 million, and more than double again by 2015 to over 108 million, or roughly one-third of the U.S. population. Shopping conversion rates are four to five percent on tablets compared tothree percent on desktops — they were nearly double compared to all mobile devices during the holiday — and order totals are 10 to 20 percent higher for tablet users. 14This is the tooltip content.
For retailers — and indeed for just about any business — this means the only truly competitive strategy is a three-screen strategy: desktop, smartphone and tablet.And if mobile (smartphone) is experiencing growing pains, on tablets it’s even worse.
“The performance on tablets is just really, really poor, in some cases, two times worse than mobile,” Rudger says. “Retailers don’t seem to know what to do with it. They usually will offer the mobile version, but customers don’t want to default to the mobile version. Or they’ll push the regular full-featured site, and the tablet basically chokes on that. So whereas on the desktop a site may take two or four seconds to load, on the tablet it might be 12.”
Few retailers sorted out a tablet strategy for this retail season, so it may not have caused much financial pain. Next year, though, with more than 50 million well-heeled consumers wielding iPads, Web-enabled Kindles and other tablets, ignoring the unique requirements of “the third screen” will risk significant bottom-line peril.
Conclusion: 2012 starts now.
Economists and retail analysts aren’t breaking out the party hats based on Thanksgiving holiday performance. Many are worried that consumers still hurting from the bad economy will not spend more this season, but are merely shifting their spending earlier to take advantage of aggressive Thanksgiving weekend deals. The good news, though, is that retailers online were able to reap the benefits of record early-season spending without major failures.
“The lesson of this year is that the Web has gotten to the point where, at least for the top retailers we’re following, they’re doing what they ought to do, and it resulted in a pretty stable season from a desktop perspective,” Karow concludes. “But we’re still seeing significant mobile pain points. There are bright spots on the mobile side, but there are definitely still weaknesses. The takeaway is that now more than ever, people need to pretest their mobile strategies, and monitor their performance over the air…and stay diligent and do load testing every time they change their site. The cautionary tale is that, any time you make changes to your architecture or even your content, you could face a critical failing point in terms of your ability to scale, both on desktop and mobile. You’ve got to have a regular regimen of load testing to prove that what you’ve done hasn’t broken things.”
“The best advice for 2012,” Rudger adds, “Start now.”