ideas, Marketing

Test, test and test again

Written by Jason Groom
  • Twitter
  • Facebook
  • Mail
  • Linkedin
Originally published
September 16 2015,
September 26 2020

It's the mantra we all chant but, not having time to test all the elements of their campaign is a common gripe for marketers. In fact, it’s right up there at number 2 in the Econsultancy “pains” hot list. Short on time the temptation can be to rely on gut instinct in order to get a campaign out of the door, but not testing can prove to be the falsest of false economies.

We're not immune either. On two occasions recently we made some assumptions (and we all know what assumptions make!) only to be proved entirely wrong. The first, a design for a direct mail piece aimed at a specific job title within a determined vertical market. Our creative erred on the feminine side, the data when interrogated though most definitely didn’t. Cue a hastily revised colour palette and imagery.

Same project and fast-forward a week. A/B testing the subject of the email to a 10% sample of the 3000 records showed, overwhelmingly, the one we’d have plumped for would only have earned us the Bronze. In fact it could have lost us a significant number of precious opens had we ploughed ahead with what we believed to be a clever, and winning, one.

So, with email messaging the prime candidate for optimising, what should we be testing, and why?

  1. Calls-to-action
    CTAs when placed as buttons or within subject lines are a must test to see what works best! Size, color, placement & tone of CTA combinations can be tested to pick the best one. Headline and copy – positive, decisive verbs are always the most successful but which one(s)?
  2. Personalisation
    You don’t want Mrs. Pitt to receive an email addressed to Ms. Aniston which is the name of Mr. Pitt’s ex-wife! Personalisation gone wrong is deadly, DO TEST!
  3. Subject line
    Analyse the length, tone, personalisation, offer and how they affect the results.
  4. From Address
    Does your data set prefer emails from a person's name, or a company name? A real person, or the brand? Test it and, if necessary, segment accordingly.
  5. Design & Layout
    Test email design against addresses spread across different email clients! You don’t want the shoes displayed below the fold in the email, if that’s the product you are trying to sell!
  6. Time & Day of the send
    If Mon, 4 PM gives you better email opens than Mon, 10 AM –Why not go for it, but you’ll only know if you test!
  7. The offer itself
    Our assumption may be that customers/prospects prefer a money-off or discount voucher. In reality they may be far more interested in a piece of content or an invitation to an event. When we can’t ask them outright and in person, testing and monitoring their actions is the most accurate way to learn
  8. Email Content
    The sales copy used in the email shall determine the CTR, experiment with various copy variations & test the effectiveness of each one!
  9. Speak the language
    Use dynamic content to find out if customers become more engaged as a result of increasingly relevant messages.

Rich vs. Plain text and other great ideas can be found on the Email Monks website

Another great tip from Email Monks is their “7 steps that might help you with winning email test results!”

  1. Keep the test simple!
    …focus on variables that yield the bigger and more important results.
  2. Keep your test campaigns consistent!
    …see to it that you compare apples to apples.
  3. Determine your end goals!
    …in terms of any email metrics, etc. & avoid having multiple goals.
  4. Check the pre-testing elements!
    …evaluate campaigns from the past to analyze what worked best and what didn’t!
  5. Frame the test!
    …set up one test for each of the elements that you have decided to test.
  6. Split the List!
    ...check if the list is large enough for an A/B split or needs to be broken into several smaller test segments. Make sure your test segments are large enough to reach statistical relevance.
  7. Change the elements!
    …test after replacing poor result yielding elements with better actionable items.

Thinking that we know best is hubris and can pave the way to potentially expensive mistakes. Being proven wrong, or even in some cases right, is a much smaller price to pay!

P.S You can apply most of the above to your Direct Mail.

Related Articles

Top 5 Remote Working Tips for Small Businesses

Top 5 Remote Working Tips for Small Businesses

Let’s face it. Even before the coronavirus pandemic and ensuing lockdown, remote working was on...

Our responsibility to help is also a responsibility to do it right.

Our responsibility to help is also a responsibility to do it right.

I think one of the things people are finding the hardest about the coronavirus pandemic is the...

Week 1: My lockdown diary

Week 1: My lockdown diary

Two weeks ago everything felt pretty normal, yes, it was Friday the 13th, but the World was...

Things to do with your kids indoors.

Things to do with your kids indoors.

If you're like me, you're feeling overwhelmed and struggling to stay aware of the...