By Rebecca Sharpe, Marketing Sciences |
As you may have heard by now, Cyber Monday 2015 was the biggest online sales day in history1. To get a better understanding of how cross-device contributed to this phenomenal success, Atlas used people-based measurement to track online shopping by device from Thanksgiving through Cyber Monday. The results of this study are included in our downloadable 2015 holiday shopping infographic.
Fifty percent of all online purchases during the long weekend happened on mobile2, a sure sign that holiday shopping is no longer confined by in-store or desktop — and a reminder of how important cross-device measurement is for modern advertisers.
As expected, we observed a spike in US online purchases over the holiday weekend. Online sales volume was higher than the recent daily average — up 20% for the entire weekend and 65% on Cyber Monday3. Contrary to the popular belief that people wait until Black Friday to buy, Thanksgiving online sales were also up slightly (4%). Only Saturday saw a decline in online sales volume — perhaps due to a shift to in-store purchasing.
Since Atlas measures the behavior of real people, not just cookies, we’re able to depict purchasing by device as well. In total, online purchases during this period migrated toward mobile — a full 50%, compared to just 41% in the week prior4.
Even more interesting, however, was device usage throughout the weekend. While on a typical day, 41% of purchases take place online, the number jumped to 58% on Thanksgiving day5. This trend extended to the rest of the weekend proper, though by Cyber Monday people were back to purchasing on desktop.
As you can see below, Atlas uncovered 33% more conversions than traditional cookie-based measurement6. Atlas also found that these purchases were influenced by online ads, even if that ad exposure was on a different device type (19%), a different browser (7%) or as part of an all-mobile purchase path (6%). Confirming mobile’s popularity for holiday shopping, the biggest chunk of this increased detection came from cross-device.
Even without the Thanksgiving holiday, Black Friday and Cyber Monday sales have taken hold in the UK too — so we measured cross-device purchases there as well. As expected, sales patterns were different from the US. UK shoppers were more likely to start their online holiday shopping on Thursday — a 29% increase in volume over the recent daily average, outpacing even Cyber Monday’s relative increase of 19%7.
Similarly, mobile purchase share was not impacted by the Thanksgiving holiday, and was in fact highest on Saturday and Sunday:
But just as we observed in the US market, the value of cross-device measurement in the UK still rings loud and clear. Atlas’ people-based measurement uncovered a 29% increase in online conversions compared to traditional cookie-based measurement — a boost largely attributable to cross-device conversions (17%)8.
Ultimately, 35% of UK online purchases over the long weekend happened on mobile9. While that number was lower than what we saw in the US, people-based measurement uncovered similar increases in attributable conversions in both markets.
Online shopping isn’t just about desktop anymore. In this increasingly mobile, cross-device world, it’s important for marketers to measure their ROI across every channel so they can make the right investment decisions at the times that matter most.
1 Adobe 2015 Cyber Monday online shopping data, Nov 30 2015. http://wwwimages.adobe.com/content/dam/Adobe/en/news-room/pdfs/201511/113015AdobeDataCyberMondaySales.pdf
2 Atlas internal data.
3 Atlas internal data. Average volume for Nov 19-25, 2015.
4 Atlas internal data.
5 Atlas internal data.
6 Atlas internal data. Conversion defined as an online purchase preceded by an online ad delivered to that purchaser within 28 days of the purchase event.
7 Atlas internal data. Average volume for Nov 19-25, 2015.
8 Atlas internal data. Conversion defined as an online purchase preceded by an online ad delivered to that purchaser within 28 days of the purchase event.
9 Atlas internal data.