As the results of some discussions I’ve had recently, I’ve been thinking a lot about coincidence, causation, continuity, and effective e-mail marketing:
“Your e-mail open and click rates were lower last month then they were a year ago. Your total send quantity was also down. Therefore you need to send a higher quantity of e-mail this month in order to get your open and click rates back up.” – ESP Account Representative
It’s always a concern when open and click-through rates decline. And you do want to investigate why, but it’s important not to jump to conclusions.
Yes, open and click-through rates were down year-over-year. Yes, the send quantity was down year-over-year. But is it causation or coincidence?
The account representative from the client’s ESP saw causation. Sometimes the obvious answer is the right one, but sometimes it’s not.
At a basic level, while increasing the send quantity would increase the raw number of opens and clicks, the percentages won’t necessarily increase, since the denominator (the send quantity) would also increase. This is just Math 101.
Going deeper, opens and click-throughs year-over-year had been down for a while; this was true whether send quantities were higher, lower, or about the same. So the hypothesis that sending higher quantities will increase performance isn’t supported when data from additional months is factored in.
Moral of the story: don’t be too quick to assume causation, make sure you do some research to eliminate the possibility that the results are a coincidence.
“The e-mail creative doesn’t have much effect on results – people either buy or they don’t.” – Product Manager
Causation is a fundamental cornerstone of direct marketing – the belief that strategy and tactics can influence action. It’s what makes marketing fun and challenging.
The product manager that didn’t believe creative had an impact on results was using this rationale to change the creative at will. When results varied from send to send, the manager attributed this to the fickleness of the audience. When we implemented a structured testing plan, where creative wasn’t changed on whim, we were able to see more stable results. Causation became clear.
Denying causation in marketing is short-sided. Creative execution does make a difference in performance, plain and simple. And the way to see this is to follow a structured testing plan. By testing any creative changes against the control, you can see which changes improve performance and which do not.
“It’s been a year since we last promoted this product via e-mail, so we need new creative.” – Marketing Manager
Another fundamental of direct marketing is the idea of testing for continual improvement. I learned from some very strict direct marketers; you didn’t change anything in your control without testing it and proving, beyond a doubt, that the change improved response.
Which is why I am so surprised when clients, like the one quoted above, choose to go with something new rather than something that’s proven, without any testing.
Here, the old creative had a very high ROI, nearly $30. While I’m all for testing a new creative against this, the marketing manager didn’t want to send the old creative at all. This is short-sided. You’ve got a proven performer; it’s a year old, but still. Continuity is important for ongoing improvement.
In this instance, we compromised. We sent the new creative as the control and the old control as a “test” creative (totally backwards, but it will still give us the comparison we need to determine which performs better). I’ll let you know the results.
Until next time,
Properly implemented DMARC should not affect your deliverability. You can guess what I’m going to say next. Last month I wrote about ... read more
Graze, the snack company which provides nutritious nibbles in slim cardboard subscription boxes, has become a regular fixture in offices, homes and ... read more
Inboxes are so crowded, how can a marketer stand out? Here are eight brands that cut through the noise with great emails. Also, we are all about alliteration.
In theory, having no DMARC record should have no impact on deliverability, but not everyone got that memo.