Maximizing Personalization ROI in Marketing: Beyond Split Testing
Many months ago, before I got pulled into a very long and rewarding software development project, we covered the day to day of content personalization in email and SMS marketing campaigns. Click here for a quick refresher. I certainly had to.
Personalization: The Long-Term Winner
With an understanding of how we will measure the “personalizationess” of content and a firm grasp on the idea that we don’t need to personalize for everyone, we just need one more high-level thought exercise before we start clicking buttons. How will success be measured? What’s a win here? We talked earlier about the value of outgoing messages; this might be a good time to refine that idea further.
Measuring the success of content personalization in Email and SMS
Reminder: revenue only represents about 40% of the total value of an outgoing message. We can just get this out of the way because everyone knows this. Make sure you get every single piece of this 40% because, as the pot grows over time, this scales up. As your personalization efforts mature, this percentage will naturally increase. Your needs for data will be less urgent, allowing you to maximize this channel (bigger paychecks).
60% = data gathered from interaction. Whether you collect the data or not, it’s here. Most of the message value for email and SMS lies in the learning opportunity presented by sending that message. Make no mistake: every user action (or inaction) is a data point. And these data points are funny; they stack quickly (compound interest).
Scenario Study, The CCCC Marketing Structure
Let’s run through an example using the CCCC marketing structure we invented in our earlier blog post.
Aquitistion is largely paid ads (Tik-Tok, Meta, Google search). The brand drives traffic to different product collections based on customer type in order to reduce friction to checkout. Our high-priced audience sees ads that are aligned with their interests, and are driven to landing pages and collections pages that support this overall experience. An exit intent popup is used to capture email and SMS opt in from users, with the creative matching the larger experience, retaining immersion in the luxury experience.
Now it’s our time to shine; the welcome series is ready to rock. Lucky for us, we already know a few things about the recipient of this particular message. We know what FaceBook ad they clicked, the landing page they hit, and the popup they opted in from. At this point, our goal is clear:
#1 Convert at the highest rate possible by leveraging the limited info we already have about the recipient. Get better at this over time as #2 evolves.
#2 Learn more about this user to create more personalized experiences moving forward. What would you do if you saw a customer standing in your store staring at a product? You’d learn more about them. So let’s do it. Use a two-way message to ask a direct question, leverage email links to provide insight into user behavior, add a drop-down on your popup giving the user an opportunity to customize their experience, create on-site forms … the list goes on. The point is to be interested in this piece of the puzzle and evolve your ability to gather meaningful insights from everything you send.
Data Collection Strategies
In a scenario where we successfully execute a welcome series, you can potentially learn the following about a new opt-in:
Ad clicked (creative/copy preference in relation to a conversion event).
Opt-in source (depending on how specific your pop-up is to this audience member, you can gain insights into product preference, interest level, or even a deeper understanding of the data from #1 (for example, a user clicks an ad for a $90+ chair sock product and then opts-in through a popup on a landing page specific to the pain point of damage to expensive floors.)
Opt-in source refinement (by augmenting your opt-in source with a dropdown or radio button, you provide an opportunity for further data gathering. For example, by asking for number of chairs needing socks with a dropdown menu, we gain additional insight into potential order value.)
Interaction with a two-way message (asking a potential customer directly what they are looking for and then simply getting it to them as quickly as possible) is an incredible way to increase conversions while learning. For example, we could provide our shopper with three-four style options,send them to a supporting landing page based on their answer, and, of course, log the answer for our future communication.
I’ll stop the example here because I’m tired of investing time into this sock company, but you get the point. If you don’t get the point, please email me at hello@jasondonapel.com and we can talk more about it.
There you have it. You’ve just deposited into your 401K. Your future you says, “Thanks, bro!”
Why Split Testing Alone Isn't Enough During a Depression
Here’s the thing about split testing your way to success: it can work. It has worked. Kind of. Rolling out tests that provide lifts is always awesome. We should do it forever. The issue is that it’s not a reliable lift, so the time investment is a risky one. What if you just don’t walk away with something actionable? How much time is gone?
When larger events outside your control shake up the industry, a Meta algorithm or Apple data update occurs—you don’t want to be testing your way to success. Short-term investments aren’t great during a recession. You know what gets you through the tough times? Yup, that’s right! All of your data; ensuring that every single message you deliver to your most valuable asset is as useful as it can be. Not just for you and your business, but for the audience members on the other side.
Alright, enough theory; let’s get to work. Go launch a popup and let me know how it goes.
