As the dust settles from the frantic Christmas and New Year sales trading season I’m now able to provide a definitive account of how our advertisers’ affiliate campaigns performed throughout December.
Having previously looked at the impact of Cyber and Manic Monday on the affiliate channel we can trace December’s performance, compare it with 2010’s and also split out mobile data to see how it compared with traditional desktop transactions.
Stats from Hitwise indicated that Boxing Day was the biggest online shopping day in terms of traffic so it is also possible to see if that was reflected across the affiliate channel.
Firstly, it is interesting to take a look at the sales revenue. In my previous article, the largest day for our advertisers in terms of sales revenue was Tuesday December 6th.
This fell the day after ‘Manic Monday’. Having looked at the remainder of December, the 6th turned out to be the biggest day of the whole month with £7m of revenue delivered for our advertisers.
If we look at the remainder of December, the lowest day in terms of sales revenue was Christmas Day with just over £2m in sales revenue generated. The biggest day in terms of revenue post Christmas was 29th December with sales topping £5m.
Despite Boxing Day being stated as the largest online shopping day by Hitwise, this was actually one of the lowest for Affiliate Window when it came to revenue. However, it was double that recorded on Christmas Day.
If we look at how this compared to Christmas 2010 there were very similar trends (please note, we have compared the beginning and end of the month):
The peak that was seen on the 6th December was also apparent in 2010 (where the 6th actually did fall on ‘Manic Monday’).
Again, Christmas Day was the lowest in terms of sales revenue in 2010 and 2011. The main differences that we see are in the post Christmas period where there are different peaks and troughs for each year.
We can now look at the actual number of transactions. The 6th December was again the largest – not surprising considering that most revenue was generated on this date too. Amazon predicted, and later confirmed, that December 5th was their largest day, primarily driven by sales of the Kindle.
As with sales revenue, Christmas Day was the lowest in terms of number of transactions (second lowest for mobile). Again, the 29th was the largest day for post-Christmas sales (also through mobile devices).
Pre-Christmas sales via mobile peaked later than desktop, hitting their stride on Sunday 11th December, possibly reflecting how consumers interact differently to desktop but also showing how mobile peaks at the weekends, (also reflected in the growing ‘Mobile Sunday’ phenomenon and Sunday 4th was also a strong trading day for the network); an interesting future trend to monitor.
Also intriguing is the resilience of mobile sales post-Christmas. A combination of significant new handset activations, tablets given as presents and less deskbound consumers surely helped mitigate the post-Christmas dip.
(Mobile sales plotted on a second axis)
If we compare the number of transactions across the whole network to 2010, including all mobile devices and desktop sales, the trends are also very similar with 2011 sales mirroring those of twelve months earlier. 2011 had an additional peak on the 2nd December which was not seen in 2010.
Looking at traffic (clicks), the key trading Mondays received the most amount of traffic alongside Tuesday 6th December. However, if we look at mobile traffic, the peaks came a lot later in the month. Each of the peak days for traffic through mobile came after Christmas.
Incidentally, Boxing Day was the highest through mobile. In conjunction with our earlier data on transactions the trend browsing rather than transacting trend is clear as conversion rates fall away.
Also worth considering is whether the heightened awareness of specific brand sales resulted in consumers transacting directly with those retailers, cutting out the affiliate middleman.
This could also account for the similar trends we saw with cashback and voucher code sites, with consumers assuming that due to sale period they were receiving the best deal by transacting directly.
Traffic through desktop for Boxing Day was again low (only the 19th highest day in terms of traffic). Hitwise data is based on traffic rather than transactions; however our data still shows traffic through the affiliate channel to be low on Boxing Day.
Finally, we are able to look at the average order value per transaction during this period. More has been spent on average through the affiliate channel after Christmas than prior to it. The top seven days for the highest AOV all came after Christmas Day and were all over £70.
Prior to Christmas, most of the AOVs were under £60. This indicates that post-Christmas sales encouraged visitors to spend more.
If we compare this to Christmas Day 2010, again, a relatively similar pattern is followed. However, AOVs were a fair bit higher in 2010 on each of the days.
There was also a major peak in AOV before VAT was increased in January 2011 with customers taking advantage of offers before this increase. This was especially the case across the electronics sector, a trend that was never likely to be repeated.
So what conclusions can we draw from all this data?
In essence it’s a confusing picture and one that is difficult to generalise about.
Having spoken to some of our key affiliate partners we know their traffic and trends broadly followed ours but with significant discrepancies on a specific advertiser level.
We also looked at half a dozen of our big, blue chip retail clients and picked peak sales and traffic days for them. Not one of them matched with another, all producing wildly differing results, ultimately proving the influence an advertiser can still have on shaping consumer behaviour.
One of the most fascinating and clear cut patterns was the spike in post-Christmas average order values. An equally interesting task would be to look at which products over or under indexed before and after Christmas.
Finally December saw a significant spike in mobile traffic, setting 2012 up to be the year when affiliate traffic through mobile devices could hit double figures for the first time.
It’s impossible to say at present when that is likely to be but there’s no doubt the affiliate channel continues to throw up fascinating retail trends for data geeks everywhere.