In order to track ROI, your advertising must be measurable. How many orders did a marketing activity produce? How many clicks, calls, shares, visits, responses, or in store visits did it generate?
All this must be measured to determine ROI among other things like what works creatively to inspire interest, desire, and ultimately, action.
If you are currently conducting any type of marketing to generate inquiries and sales you already have something that works and can be measured. For example, if you send an email message and see that 50 people clicked a link to your website and ordered, that is the number you need to beat with your great new idea. This would be your baseline.
But instead of creating a new email to test against your last email, which won’t allow you to determine exactly what components are working for you, try A/B testing.
A/B testing, or split testing, compares two versions of a similar marketing piece — a print ad, an email message, direct mail package, PPC ad, landing page — to determine which may work better with different elements.
In the case of an email message you have several elements to compare:
- Subject Line
- From Name
- Headline
- Imagery
- Copy
- Offer
- Call to Action
In an A/B test we set up two variations of the message, deploy the message to a segment of the email list, and analyze:
- Open Rate
- Clickthrough Rate or Response
- Number of Conversions
Let’s say we’re selling stuffed teddy bears. Below are examples of various elements to test:
Subject Line Version A: The Softest Stuffed Teddy Bears You Will Ever Hold
Subject Line Version B: You Never Knew A Teddy Bear Could Be This Soft!
From Line Version A: Plush Toys
From Line Version B: Amanda at Plush Toys
Offer Version A: Order Now and take 15% Off!
Offer Version B: Free Shipping!
Image Version A: Photograph of the Stuffed Teddy Bear
ImageVersion B: Photograph of a Child Holding the Stuffed Teddy Bear
Now if you change all of these elements at once, you would not fully know which element accounted for an increase in responses or sales. With a true A/B test, you keep all elements the same, but one, i.e., Test Subject Line vs. Subject Line. Take the winner of that and test From Name vs. From Name. Then take the winning combo of Subject Line and From Name and test Offer vs. Offer, Image vs. Image and so on.
In this way you can be sure what is working and craft creative accordingly going forward.
There is never an end to the testing you can conduct and it can get tedious, but by doing so you will amass a body of evidence to support creative direction, advertising budgets, and even product development for future campaigns.
Comments