All posts in A/B Testing

Google Image Ads AB Test – ‘Buy Now’ versus ‘Order Now’

google ads ab test

Understanding human behaviour (online and off) has always been an area of huge interest for me. AB testing is a great way to identify why some people do what they do…online at least.

What is an AB test?

An AB test is a test where you run two ad variations. To make the effects measurable you run two almost identical ads with one changing variable.

In the case of this test we are comparing the effects of using the call-to-actions (CTA) ‘Buy Now’ and ‘Order Now’.

What was the AB Test this time?

The idea of this test was to see if more people clicked on an image ad which read ‘Buy Now’ or ‘Oder Now’. Both variables are verbs, both instill urgency so which one would you guess is more successful?

If there’s one thing that running these tests have thought me it’s that you can never predict peoples behaviour  – you ALWAYS need to test.

Where did you run this AB test?

For this test I was displaying two almost identical ads on Google’s ad platform – Google Adwords.

The images below are just one size variation but I did upload 9 diffrerent variations for each campaign.

The ad was run in the weeks leading up to Christmas 2011, hence the theme.

What did the ads look like?

Google ads AB Test

Buy Now version

google ads ab test

Order Now version

 

So, what were the test results?

The image below is a snippet from my adwords campaign:

 

 

google ad ab test results

 

From this we can see that both ads were ranked excellently  (an adverage position of  1).

The CTR (click through rate, number of times the ad was clicked on for the amount it was shown) is quite low, 0.08% and 0.16%. This is natural enough as these campaigns were displayed across Google’s display network.

We also enabled Google Remarketing, this is a practice where our ad follow you online if you visit our site…I still have mixed opinions on this but I’ll run with it for the moment.

 

So who’s the winner of the AB test?

The ‘Buy Now’ ad was displayed slightly less (270,441 times) but was clicked on TWICE as much! The only difference that caused double the clicks was the verb on the ad, ‘Buy Now’.

What’s also quite interesting is that Google had a current adverage price on this ad which was 21% cheaper. The money is a bit irrellivent in this example as it’s small but imagine the difference on a large campaign!

Email Marketing A/B Test 2 – Content Image

Email Marketing A/B Test 2 – Content Image

So this is the second of these tests that I’ve decided to run and, on this one. I’ve included an image of me looking a bit goofy on a bike (and a lot balder looking than my internal image remembered myself!!). I’ve intentionally put this image above the page fold so it made the most impact.

For people whom are unaware, these A/B tests runs over 4 hours and the two test groups are identical in size groups, all randomly selected from our email subscriber list. Each group receives a different version, one A, the other B. I use Campaign Monitor for these emails, there product offering is great. If you’re considering emailing customers it is THE way to do it.

What this test is measuring is the % that recipients open the email. The difference isn’t massive in this case but 6% is still nothing to be sniffed at!

A slight difference in open rate, not massive but still a difference.

Version A had the image, Version B had none. The result confirmed what I presumed: that people are more likely to open an email that looks somewhat personal. If you’re curious and want to see the emails click here for A or here for B.

One regret I have from this test is not testing with a generic, stock image, to see if the results of the stock image mirrored the version with no image. My gut feeling is that the stock image would have a similar open rate to having no image….maybe I’ll test it at a later stage.

Another thing I noticed recently is the amount that iPhone is growing as an email reader. For this campaign it was the single biggest email client – impressive stat.

Derek Traynor

The growth of iPhone shows the massive importance of having these emails working perfect for these devices.

 

Email Marketing A/B Test 1 – Subject Line

Email Marketing A/B Test 1 – Subject Line

For those out there interested in findings from email marketing A/B tests here’s the findings from an email I sent out yesterday.

Subject line A = Below Cost Clearance Of Some Allmoto Stock.

Subject line B = Below cost clearance of some Allmoto stock.

Interesting thing is that I would have swore that fully capitalised would have a higher open rate – who knew eh? t just shows that the small things really make a big difference.

It has me wondering if I should be changing all my Google Ads as they are all capitalised on the back of advice – I just hope it was tested advice:)

For people whom are unaware, these A/B tests runs over 4 hours and the two test groups are identical in size groups, all randomly selected from our email subscriber list. Each group receives a different version, one A, the other B. I use Campaign Monitor for these emails, there product offering is great. If you’re considering emailing customers it is THE way to do it.