Posted by Chris Costello, Senior Manager, Digital Research
First, as this is my first blog byline, an introduction. My name is Chris Costello, known in these parts as “Coz” due to a proliferation of employees named Chris when I started at Performics, and I head up the research team at the company. If you’re wondering who slipped two Van Halen references into a search trend report that addressed Yahoo’s “Panama” platform a couple of years ago, that was me.
Second, we just released some data on our retail clients’ performance on Black Friday and Cyber Monday, and I figured I’d take this opportunity to add some context around it. The big headline, and the part that actually surprised me quite a bit, is that Black Friday outperformed the “first” Cyber Monday. We pulled data going back six years, and this is the first time that’s ever happened. As if that weren’t enough, that first Cyber Monday – we’re on record as saying every Monday between Thanksgiving and Christmas is a Cyber Monday – also grew substantially over last year. Forty-three percent, to be exact.
The first obvious question is, forty-three percent of what? The metric we’re using here is “Sales per Advertiser,” which is total sales divided by total active advertisers. We set a minimum threshold of daily sales, so as to filter out some noise at the bottom of the range. In past years, we added one more level of complexity and used “Sales per Relationship,” which considered each advertiser-search engine partnership separately, but this could end up punishing the overall average as advertisers build their programs beyond the big guns in the search space. Plus we left out most of the smaller engines this time around.
Dividing by the number of advertisers helps alleviate any distortions that may be caused by year-over-year changes in the number of clients on our roster. In an ideal world – and the folks over at RKG took this approach to their data – we would just look at advertisers that have consistent data, but with the six-year time window, that breaks down rather quickly. Obviously, there’s some variability introduced by the large vs. small distribution of the client roster, but I prefer to include more advertisers in the analysis, and if you squint really hard, you can look at changes in the client roster of a big SEM agency as indicative of changes in who is doing paid search. So I can live with it.
The second question is, what just happened?
There are two possible causes for this big change in behavior. Either advertisers did something different to get people to their websites via search, or consumers collectively acted differently than they have in the past. As is usually the case, it’s probably both. If you consider the mindset of the advertiser, you’ve got to get skittish consumers to buy, and you’ve already got a lot of energy directed toward your offline efforts for Black Friday. Do you spend time, effort and money crafting a separate campaign for a separate day for a different medium, or do you just extend what you’re already doing offline to online as well? With ad budgets getting tighter, is it fair to even ask that question?
From the consumer side, there’s a larger picture. On the one hand, all the talk of cutting back may have kept more people at home on Black Friday, but on the other, they just couldn’t kick the habit of shopping. Coupled with the enticement of the Black Friday deals, this set the table for a really strong day.
Slicing and dicing the data yields one more important nugget. Bigger-ticket retailers definitely did a lot of the heavy lifting on Black Friday. I haven’t gone so far as to stratify the results by average order value, but a quick look at some of our bigger guns paints a picture of deep discounts working very well for these advertisers. This definitely goosed the sales per advertiser numbers, and even drove up average order value overall for the week of Thanksgiving. Within search, higher-priced products have been where the effects of the economy have been most noticeable, so this sort of aggressive approach to that segment of the market seems to have done the trick.
Finally, the compressed holiday season may factor pretty heavily in this shift in behavior. You could argue that the recession might make consumers wait until Thanksgiving to start their Christmas shopping in earnest, either in anticipation of the deals or purely to manage cash flow. Once we crossed Thanksgiving, though, the realization that Christmas is only four weeks away might subsequently cause consumers to “front-load” their holiday shopping. We’ve seen year-over-year growth numbers by day and by week that are pretty healthy when you time-align the data to Thanksgiving, but the big unknown here is how long that year-over-year growth will be sustained throughout the season. We haven’t seen anything like this before, so it’s tough to predict what comes next. My gut is that it will start to tail off either the middle of this week or next week, but the early growth should be enough to support the low double-digit growth that several analysts were predicting for the overall season, even with that season being a week shorter than last year.