4 minute read
Personalization is a major goal for most email marketers. But one of the factors holding them back is not having the good-quality data they need to do it effectively.
In “Email Strategies for Success,” a 2017 research report by the email service provider Upland Adestra, 77% of respondents said they believe email marketing effectiveness is increasing significantly. Also, 64% said improving email personalization was one of their most important goals, but 51% of the respondents said that data quality is the biggest barrier to personalizing effectively.
As this study makes clear, executing “personalization” is not as easy as everyone makes it out to be. It all boils down to having “good data.”
In 2020, I am willing to bet that the need to implement personalization remains as strong as ever, but marketers are still having the same issues with data quality.
Below I will highlight two of the important steps needed to unlock the value of email personalization.
Email program owners must agree on what and how they define a personalized email program because that will ultimately drive how/what types of data drive it. Each organization should have its own definition and not be too concerned about how people not close to the program define it.
To some, personalization will include purchase or browse data and integrating psychographic and demographic data points with predictive analytics, driven by machine learning to serve up unique content for each customer .
Others might define personalization with location-based messaging, first-name personalization and perhaps a birthday email.
Let me be clear: There is no right or wrong way to do personalization in email. Although most organizations say they want their messaging to be “more personalized,” it starts with definition and then knowing what you have and the level of effort it would take to pull off an effective personalization strategy.
It’s easy to fall into the “wishful thinking” trap of personalization, where you talk about it but you never take the steps required to start doing it.
After you define personalization, you need to figure out what data you have, how accurate it is and how you are going to use it. Begin by auditing all of your data points and assigning a level of importance on each piece of data that should have the greatest impact on the program.
For example, you might have gender data on 75% of your subscribers. But if you don’t segment and target your messages by gender, it’s useless.
Some organizations might find that they have a certain kind of data on only 10% of their subscribers. That doesn’t mean that data is statistically insignificant. A closer look at the value that those subscribers deliver for your program might prove that data point useful in providing a more personalized experience.
A good place to start your personalization effort is by gathering the big data points you have on your subscribers and using them to test for impact. You might have to start with “first name,” but don’t limit it to just testing subject lines with or without the first name. Think outside the box and test it in places inside your email where you know have high degrees of clicks or heat mapping.
Or, test whether putting the first name in the preheader gets more opens than putting it in the subject line. You might be surprised! You have to test your way into effective personalization, even at this basic level.
The next step is to understand the next level of data you have and how you can use it to your advantage.
Perhaps you are a publisher with multiple types of newsletters or categories inside your program and use predictive modeling to recommend other content or newsletters for greater reach. This has long been thought of the “we thought you would like” or “selected for you” types of messaging inside email. However, my experience has always been that those types of messages are without merit or intelligence behind them. Don’t fake it, especially in today’s environment.
The next steps you take will depend on your brand, your email and personalization goals, testing capabilities and data resources. At Trendline, we use all of this data, and more, to develop individual personalization plans for our clients.
The bottom line: Rule nothing out until you weigh the impact of using the data.
This is the biggest hurdle to overcome in the process. You have to get out of your comfort zone, reach beyond your guesses and test, test again, and then test some more. It will never be perfect, and you might find that data you once thought of as valuable turns out to be worthless. But you might also find data you thought was worthless might actually be good. You won’t find out until you test.
Let me be clear about testing. You don’t want to haul off and test some monster data set all at once. Ease into testing, and do it in small batches to determine if it’s right and if it works.
To do this, your organization needs to cultivate a “test and learn” culture that tolerates failure. You also need a solid testing plan and the patience to learn because you will need time – maybe months – to present clear evidence.
The path to personalization begins with good data, but it also involves auditing, validating and testing that data. You have to begin somewhere. That first step might be small, but can lead to big things if done effectively.
Ready to send better messages?
Sign up for our news, resources and updates. The inbox is our favorite place after all. We’ll make sure it’s worth it. (You can unsubscribe at any time, but you probably already knew that.)