Clicks, conversions and Christmas
Monday, October 13, 2008
The holiday period will soon be arriving and retailers want to know what to expect. Last year Hal Varian, Google's Chief Economist, provided us with data and analysis to better understand the trends relating to online consumer research and purchase behavior during this critical period. The information proved to be invaluable to our retail clients in preparing for the holidays and understanding the week by week holiday traffic and associated cost and click metrics. Therefore, we are once again turning to Hal to analyze the 2007 data.
The chart below shows median CPC, median conversion rate, and median CPA during the 2007 holiday season for US Google AdWords advertisers who use conversion tracking. The series have been normalized so they all start at the same point on October 1. Click on graph below.
Lots of people focus on CPC (cost per click) but ultimately the performance measure that really matters is CPA (cost per aquistion). These two metrics are tied together by the conversion rate:
CPA = (cost/click)/(conversions/click) = CPC/CR.The chart below shows median CPC, median conversion rate, and median CPA during the 2007 holiday season for US Google AdWords advertisers who use conversion tracking. The series have been normalized so they all start at the same point on October 1. Click on graph below.
The most obvious feature in the plot is the dramatic increase in the conversion rate (green line) that starts around December 1 and continues through December 17, the peak online shopping period. During this period, the median cost-per-click (red) shows a small increase, but the conversion rate (green) goes up by even more. The result is a significant drop in CPA. So even though clicks are slightly more expensive during the holiday buying season than during other times, the cost per conversion is much lower.
It is also of interest to look at how the number of clicks changes during the holiday season. Click on graph below
We see a steady march upwards in clicks starting Oct 1, with two noteworthy peaks: the first on Black Friday, the day after Thanksgiving, and the second on December 10, which was the peak of the online holiday shopping period in 2007. We may see something a bit different this year, since there are 5 fewer shopping days between Thanksgiving and Christmas. Furthermore, given the current economic situation, consumers are likely to be hunting for bargains, so holiday sales of excess inventory may well kick in earlier than normal. Both of these facts suggest that merchants should get an early start on planning their 2008 holiday ad campaigns.
It seems like Google AdWords are still working. You really have to manage your program. You have to adjust for days that are traditionally slow and high. These charts help give me confidence in cranking back up my campaign.
ReplyDeleteExciting to see dedicated retail information!
ReplyDeleteThis is shaping up to be a challenging fourth quarter to navigate. I'll be interested to see your insights in the coming weeks.
Do you know if the conversion tracking includes or excludes phone sales generated by the website?
ReplyDeleteHi Eric - The conversion tracking excludes phones sales. Thanks!
ReplyDeleteYou said there are 5 less shopping days between Thanksgiving and Christmas, how so?
ReplyDeleteHi Julie - In 2007, Thanksgiving fell on November 22nd. This year, Thanksgiving falls on the 27th of November; leaving five less days for post-Thanksgiving Day shopping. Since we know that traditionally Thanksgiving is a "psychological marker" in consumers minds for "time to start holiday shopping", it is important to note that this marker falls five days later this year than last year leaving five less days for shopping compared to last year. Retailers will want to work this into their promotional strategies and encourage shopping to begin earlier.
ReplyDeleteGoogle Retail Team
We are following conversion tracking very closely and scrutinizing daily spends more than we have in the past due to the current market conditions.
ReplyDeleteHey, would you mind if I share your blog on my social media channels? There are a lot of folks that I think would enjoy reading your content. Please let me know. Thank you.
ReplyDeleteAndroid app development
iOS App Development
Freelance software developer