Testing is integral to the success of any effective marketing programme. It's what helps a brand to learn what does and doesn't work in order to better inform future communications and improve overall programme performance.
Whilst few would disagree on the importance and value in testing, there is a common misconception that it's difficult and time-consuming to carry out. If you're working with a marketing agency and have a campaign management team in place, this support should help make the process of testing more straightforward for you.
The below list includes some of the elements to consider when thinking about an effective testing plan. The points relevant to you will depend on the specifics of your marketing programme.
- Test your data. Just because you’re happy with the data you are using, it doesn’t mean that there aren’t other segments that could perform as well as or better than your current audience, so having a data test section is still a great idea
- Explore new data segments. An analyst can take your transactional data and use this to inform which other data segments are best to prioritise testing
- Think about sample size. The volume of data that you would need to test to enable your results to be statistically significant, is dependent on your expected response rates. We use a great tool to help us calculate this - you simply enter your expected response rate and you'll be provided with your required sample size
- Plan your design. When it comes to direct mail, an A5 postcard format can often be the default starting place for many brands, but if you have a vast range of products to promote, then a 6-page or 8-page A5 booklet could be worth testing and the cost isn’t significantly different
- Experiment with content. Test different content within your creative
- Think about volumes. Again, the volume you need to test to ensure that your results are statistically significant is dependent on your expected response rates, as above
- Consider your copy. Format options to test email can be pretty limited, other than length. But length of content can be a great test to run. A long email full of content and creative looks pretty and allows you to promote all your key messages and products, but sometimes a short & sweet email is all that is needed. Think of the time you could save on a short email vs. creating a longer message
3. Email Specific
Email allows you to test a whole range of things. Below are some examples of what you might test - though this is by no means an exhaustive list.
- Frequency - the size of your campaign will depend on how many weeks or months you might need to allow for testing
- ROI - we measure success based on conversion / ROI. For some client programmes, we have reduced email open rates at the same time as increasing ROI; this is where an email is targeted to those more likely to engage and go on to covert
- Subject lines – these have the potential to be most impactful and we usually separate testing by offer led vs emotive led copy, long vs short copy, and use of icons (but be aware of device compatibility when using icons)
- Pre-header text – this is often forgotten, so make sure to include this in the first place. Be mindful that on mobile devices, pre-header text displays more characters than those that can be displayed in the subject line
- Day of week – start with what you know works well for other programmes you run, or the day when peak activity happens on your website and use this as a starting point
- Time of Day – we would typically split email sends into morning, afternoon or evening as a recommended starting point & then go from there
- Content – this is brand specific, but you are likely to have lots of different types of content to experiment with and test
- Follow-ups / re-mails – does sending a reminder email a week or two after the first email help uplift your ROI? A reminder close to an offer closing date can also help to add urgency
- Innovations – there is a whole long list of email innovations that you could test, some of these include: animations, countdown timers, scratch-it, pixel art...
Test one thing at a time
Finally, remember to only test one thing at a time. How else will you know which change made a difference to your marketing programme?
Talk to us today about how to achieve better results for your next marketing programme by introducing a testing plan.