Phil Rothwell is an ecommerce expert with more than 20 years’ experience helping retailers to sell online. He was there at the beginning of ecommerce and played a leading role in growing Actinic Software from dot com start-up to floatation on the London Stock Exchange
One of my clients recently introduced me to SendGrid, a US-based email platform. SendGrid is interesting for a number of reason, the most obvious being that it is relatively low cost.
A couple of features that are proving useful are Segments and A/B Testing.
I am going to talk about Segments another time and focus on the A/B Testing for today. What is especially interesting about this is how effective and easy to use it is.
What is A/B testing?
A/B testing is a process that allows you to identify which of several different options (they don’t have to be marketing related) is likely to generate the best results.
For instance, if you have ever scratched your head wondering which of two email Subject Lines is best, then A/B testing is a tool you can use to work it out.
A/B testing achieves this by sending different versions of the same email to two randomly selected samples of email addresses and records the “open” and “click-through” rates for each version. It then compares the results so you (or SendGrid) can use the most effective option.
Don’t be fooled by the name, you can conduct A/B tests on more than two options.
Why is A/B testing important?
The problem with marketing is that it is often a speculative activity. We write content and design emails that we believe will stimulate a response from potential customers based on our experience. The problem is that we do not know what the return on our efforts will be until we see the response.
At the same time, optimising marketing content is essential because small changes to the words used in an email or its visual design, can generate very different outcomes. It may not sound like much, but a 1% improvement in click through rate achieved using a compelling Subject Line, would equate to 1000 extra clicks on a mailing list of 100,000 email addresses.
Using A/B testing you can improve the efficiency of your online marketing and make more money by objectively testing different versions of communications to find the best one.
Why don’t people use A/B testing all the time?
Mainly because it is quite difficult and time-consuming to do manually.
For many people doubling the time they spend creating promotional emails negates the benefits. Manipulating mailing list is also well outside a lot of marketers’ comfort zones, generating random samples is a nerdy activity. In fact, A/B testing is bit of a nerdy activity.
How does SendGrid solve the problem?
What I like about SendGrid’s solution is its simplicity; you can test a campaign in one of two ways.
You can either:
- Change the Subject Line of the email
- Change the content of the email
Depending on which option you choose, you either enter different Subject Lines or create multiple versions of the email with different content.
How do you run the test?
Running the test is very easy, which is what I really like about SendGrid.
Once the different version of the email has been created, SendGrid will automatically send each version to different samples of you mailing list. After a specified period of time it compares the results and, using the specified criteria (open rate or click rate) and selects the “winning version”, which it sends to the remaining people on the list.
Why do you like SendGrid’s approach?
What I like about this is that it is not rocket science. Having spent too much of my time speculating about what the best Subject Line might be for innumerable emails, I now have a tool that can do the work for me.
More importantly however, using SendGrid adds very little time to my effort. That means I can be very confident that of a positive marginal return on investment for the additional work.
Need some advice?
If you would like to start using A/B testing in your online marketing, but do not have the time. Please feel free to contact us directly at +44 (0)1932 450654 or using the form below.