To Optimize Or Abandon: Balancing Data-Driven Decisions With Intuition

One of my favorite aspects of email marketing is that we have a distinct advantage in getting performance metrics immediately when we send a campaign. I jokingly refer to it as “design by the numbers.” We get the instant gratification along with the long-time lessons associated with weeks and weeks of metrics reviews.

Our process is to review metrics and use them to inform our design, copy and marketing decisions.  In other words, use them to optimize our campaigns. With optimization we can make small improvements in an iterative process. Our creative and marketing strategies constantly evolve because we can see how our subscribers interact with the decisions we make all the time.

We’ve done a lot of testing as an industry on implicit vs. explicit preferences from our subscribers. Time and again, customers’ implicit preferences tend to favor the marketer. We find that customers will engage with three messages a month even if they “say” they only want to hear from us once a month.

We continue to make decisions based on small cues from consumers -- what they clicked on, which subject line got them to convert -- all in real time. 

While constant tweaks and improvements are the key to much of what we do, there’s an apex, in even the most successful campaigns, where you can start to see negative returns by over-optimizing. As marketers, we need to balance our data-driven decisions with our intuition. We have to keep the big picture in sight while swimming in the details.

How do you know when to optimize and when it’s time to start over? Here are two gut-check questions to ask yourself:

Question: Are you seeing a decline in your engagement metrics?

Sure, that’s an easy one to spot. But only by looking under the hood can you start to understand why it’s happening. Are all of your clicks moving from your content to your navigation? Do you see minimal differences between click-through rates, no matter the marketing message?

Tactic: Try returning to an older version of your message and run an A/B test against what you’re currently sending. This is also a prime opportunity to ask yourself if this is the right communication vehicle for your consumer. For example, rather than investing resources in making small gains by “optimizing” your newsletter template, you might accept that your subscribers don’t want a newsletter. There might be a better way to communicate this content to your subscribers to get more engagement.

In other words, by looking at the bigger picture, you might realize that abandoning your current approach and trying something new is the best way to address declining engagement.

Question: Can you easily name your top four or five customer segments?

Sometimes in our quest for optimization and personalization, we begin to create these Frankenstein templates with as many as six dynamic regions with six or seven different modules that swap in and out. There are two problems here: One, it still takes human effort (for the most part) to create all of those snippets of code, which can drain your resources. Two, the marketer has less control over ensuring all that content fits in a theme or supports the primary message -- making for fragmented customer experiences.

Tactic: Spend a little more time researching your customers and start to identify profiles they may fall into. From there you can do some light segmentation based on interest. In B2C, you want to make sure your segments don’t fall below 100,000 subscribers. It becomes too difficult to make out statistical differences when the group becomes too small. In B2B. you have to be extra careful. But use the rule, if one of the companies I’m sending to was on holiday, would it affect my metrics? Start to craft default messaging that can work for more than one segment, so it’s okay if customers receive that version of your email.

Finding the balance

We’re all doing our best to improve KPIs, drive ROI, give customers an excellent experience and still get home in time for dinner. Let’s take a step back this year and see if there are smarter and more interesting ways we can be doing our jobs.

Tags: email, metrics
Recommend (12) Print RSS