Thursday, 13 May 2010
Testing , Testing, 3,2,1!
Ultimately, testing is a key plank of any strategy to optimise your email marketing ( or indeed any form of direct marketing) so helping narrow down creative alternatives ( for example ) you have from say three to two to one ( 3,2,1..get it?).
Why is it important to test?
The simple answer to that is it helps you discover what works. With email, by testing the subject line and/or images of your email campaign, you can easily adapt the email to the preferences of your recipients. By showing recipients what they prefer to see, you increase the click-through rate and maximize the response rate of your campaign.Simples!
But I’ve worked with clients who , when I’ve suggested testing, have come back with the standard responses of ‘ we’re getting good results as it is’ and ‘ we can’t afford to test’.
So let me try and dispel these issues
‘ we can’t afford to test’
I would suggest that as an organisation you can’t afford not to test
A MarketingSherpa article by Anne Holland proved, through survey results, that testing increases ROI. The responses show that “in every case more than 50% of marketers improved ROI (even if only moderately) by testing.”. So by not testing , you are not going to maximize your ROI
‘ we’re getting good results as it is’
Which may be true at the moment, but as the world we live in changes, so does how consumers interact with the messages you send them. You can put money on the fact that your competitors will be aiming to look at what you’re doing and improve on that and get even better results. I have yet to see case where when you start to compare like for like campaigns, there isn’t some form of degradation of results over time if no test learn and refine strategy is employed.
Some other issues are seen in an article written by 8seconds , where they highlight a survey by eROI that shows that 37% of email marketers do not test their email campaigns.
The survey also highlights the main reasons behind this, these being
• I don't know how to test (32.84%)
• My campaign timeline is too short (27.36%)
• Platform doesn't have testing capabilities (13.43%)
So what should you be testing ?
First of all, lets not go crazy and try and test everything. If you are just starting off in email marketing . As I highlighted some time ago, it’s important to keep things simple and not to lose sight of the wood for the trees!
Phil Storey’s blog gives some ideas about what to Test , and as an email marketing creative it gives some great pointers.
So in terms of what you should be testing in an email, here are some ideas, in some order of increasing complexity :
• Subject Line A/B test frequently
• Broadcast time at least twice a year
• Offers/Incentive at least twice a year
• Creative test at least twice a year
• Multivariate testing of creative elements in Email templates at least twice a year
(You can find more Best Practice from eCircle)
But Multivariate Testing & Optimization allows you to test multiple variables at once, and in realtime. Essentially this approach allows you to test multiple creative options.
I’ve come across 8Seconds Optimizer that provides this Multivariate Testing & Optimization
Essentially it optimizes the images (calls to action, offers, promotions, header images, banners, buttons, etc.) in your email. 8Seconds shows different images to different recipients that open their email and measures statistically and in real time which image is the most successful in terms of clicks or conversion. 8Seconds Optimizer automatically shows this best image (or combination of images) to the next recipients that still have to open their email at that moment in time.
The first people that open the email will be the testers so that remaining people only see the optimized version. Very clever!
So what are the Benefits of Testing
Innovation – Testing allows you to understand what actually works and removes the mentality of ‘we’ve always done it this way’
Behaviour – Your email marketing will be based on solid customer behaviour as opposed to gut instinct
Speed – this can all be done real time!
ROI – it pays to test
But remember, what you optimize in January, may not be what works best 6 months later – so be prepared to go 3,2,1 all over again