A-B Testing for Direct Mail

5th Oct 2017

Our customers use direct mail for many reasons – to raise awareness of a brand or cause, to keep in touch with their members, pass on important information or updates, and of course to increase sales. Direct mail is a tried and tested channel with a proven track record in these areas. However effective a tool or service is though, there’s always room for improvement.

We’re continually looking to help our customers improve their return on investment. In order to achieve this we look at every stage of the direct mail process. The first thing we consider is data quality. We have a wide range of data cleansing services to ensure that mailing data is correct and properly formatted and we setup up our templates carefully to ensure that all variable data is well presented.

We can then help analyse the design and copy to glean the best possible advantage. Is the message relevant to the audience and is it is presented in a way that is clear and easily understood. Everything should be scrutinised, from hierarchy of layout to choice of colours and fonts. The delivery timings and size and format of the final mail-piece are also factors we should consider.

However you research and prepare before you send out a direct mail campaign it’s important to remember that since there are so many variables, other case studies and research can only be applied to your activity up to a point. The solution to this problem is to continually test, analyse your results, learn and make incremental improvements. This way you are gaining knowledge which is directly applicable to your circumstances, products and services or cause and your target audience.

What is A-B Testing in direct mail?

A-B testing is the process of creating a variable within a direct mail campaign then assessing which variant worked best, usually measured in terms of response received. Once you’ve completed the exercise you can apply what you’ve learned to future activity and try another variation next time.

Whilst it’s called A-B testing, there’s no need to restrict yourself to just 2 variations, you can try many alternatives. One of the key benefits of modern direct mail technology is the ability to easily vary the message. Using our variable data software and digital printers we can personalise almost any aspect of a mail-piece, while still retaining any available bulk postage discounts. Of course, like any experiment, it’s important to only change one element at a time and tightly control all other possible factors.

Bear in mind though, that you need to have a reasonable size sector to get reliable results. Taking a sample which is too small may give you a false reading.

How are results measured?

There are a range of options available to measure the response of a direct mail campaign. You may ask customers to redeem a voucher code. If you’re directing people to a website you can print a QR code on your mail-piece and track the responses that way. Even without marking your print you can match enquiries, purchases or registrations individually back to your mailing list. Possibly the easiest approach is to simply mail on different dates, provided you are satisfied that no other factors interfered, you may need to run the same test multiple times to corroborate your results if you use this method. All these methods can be combined together in one piece of direct mail to deliver a clearer picture.

Whichever method you use won’t capture all sales generated. Like all marketing channels, it’s not possible to precisely measure the entire return of a direct mail campaign – how much the awareness created has contributed to future activity. As long as you can identify a trend in the results you’ve achieved your objective.

So what elements should be tested?

Almost any metric of direct mail can be tested. Here are a few worth trying:-

We’d be more than happy to get together and discuss how you could integrate A-B testing in your next direct mail campaign – helping you get a better return from your marketing budget. Get in touch for more info.

Back to the news