In my previous article on Tracking Conversion Performance, I referenced an exercise that was run between October and November 2012 to understand conversion performance based on the first touch point by the visitor, either the home, registration or log-in pages.
I talked through some best practice for Adwords campaigns and how to set up essential conversion tracking on campaigns to get visibility of the conversion performance of individual ads, known as “split testing”, and individual keywords. This is vital when you are running any paid traffic campaign, and there will be more on other paid channels to market in future articles. But for now, the conversion results I’ll share with you below are based on Google PPC campaigns.
The results are naturally influenced by the keywords in the campaign and the quality of the ad copy. The cost per acquisition is driven by the cost per click, which again is influenced by keyword quality score, i.e. relevance to the site, and the maximum cost per click I have set in the campaign. The results are also aggregated across both Search and Display networks, and across all Devices.
The first test results were included in my previous article on “Tracking conversion performance”, and were for my site Women.com-SeekingMen.com. In this test, the registration page converted best from a plain text ad. Keyword insertion generated more clicks, but converted at around a third less that plain text ads, generating a higher CPA.
The next test was conducted on Men.com-SeekingWomen.com, and produced very different results to the first test. The ads were set up to land on three different pages – home, registration and log in.
Plain text ad CTR 0.25%, CPA £0.36, conversion rate 4.71%
Keyword insertion CTR 0.47%, CPA £0.32, conversion rate 7.72%
Plain text CTR 0.23%, CPA £0.37, conversion rate 4.54%
Keyword insertion CTR 0.39%, CPA £0.39, conversion rate 7.78%
Plain text ad CTR 0.25%, CPA £0.39, conversion rate 4.43%
Keyword insertion CTR 0.40%, CPA £0.47, conversion rate 6.34%
For this campaign, keyword insertion has the highest CTR and conversion rates;
The registration page as the landing page has a slightly higher conversion rate than the home page, but the CPA works out the lowest on the home page.
In summary, CPA is lower on the home page than the registration page for keyword insertion ads; Plain text has similar CTR but much higher relative CPA, as the conversion rate is around a third lower. This result is virtually the reverse of the first test!
Comparing this test with the first reveals the art of campaign management in my view – continual split testing and review of performance refines and focuses individual campaigns to what works best for the particular site. Why does Keyword insertion work much better for Men.com-SeekingWomen.com? I don’t know why, but I know that it does, so I will focus on this aspect for this campaign – armed with this information, I have paused the plain text ads to improve my overall CPA performance for this campaign, and paused the ads pointing to the log in page.
Likewise for Women.Com-SeekingMen.com, I have paused the keyword insertion ads and all those pointing at the login page, which again should improve my overall Cost Per Acquisition performance.
Onto a different niche now; Divorced Singles and a test on my site Dating4Divorcees.com, based on plain text ads only.
Text ad 1 CTR 1.34%, CPA £0.68, conversion rate 10.10%
Text ad 1 CTR 1.90%, CPA £0.38, conversion rate 17.20%
Text ad 2 CTR 1.73%, CPA £0.39, conversion rate 16.02%
Text ad 3 ` CTR 0.95%, CPA £0.39, conversion rate 15.34%
Text ad 4 CTR 0.90%, CPA £0.45, conversion rate 13.36%
Text ad 1 CTR 1.22%, CPA £0.72, conversion rate 9.09%
The registration page wins this time on both conversions and CPA, with the login page scoring the worst of the three test results. Based on this information, I have refocused the campaign to the registration page only, retaining the best performing text ads 1 and 2. I am currently testing keyword insertion ads targeting the registration page, as another example of continuous review and split testing.
Another example of where conversion tracking is essential was revealed in testing a new site in the casual dating category. I had what I thought was a great domain name, www.CasualSextasy.com, and set it up with the template below:
I set up ten separate Adwords campaigns targeting different countries so I could get visibility of each territory’s performance, all English speaking and using the same generic keyword set, and ran the campaign for a month. The overall results were, in my view, disappointing:
My conclusion from this test was that the template didn’t suit the brand name and campaign, so I paused it, changed the site template and I am currently testing this one:
I hope my two articles on conversion tracking and split testing have given you some insight into both effective Adwords campaign management, and the value that conversion tracking can give you in improving and optimising your advertising spend. Best of luck!