The 3 Big Problems with Personalization in Online Sales and Marketing
While most online retailers have a personalization technology, few know how to make effective use of it.
Personalization in the digital space is just like sex in high school. Everyone is talking about it, but few are doing it. While the technology has been accessible for years, strategy has been slow to catch up. Let’s find out why.
Unless you’ve been living under a rock for the past three years, you’ve heard that personalization is the “must-have” strategy for any serious online business. And the intent to adopt it is certainly there: According to a study by Clearhead published in December 2016, over 60 percent of online retailers have the technology to perform personalization.
So we’re all good, right? Almost everyone is getting on the personalization bandwagon and we can move on to the next big trend and get ourselves some new, shiny tools.
Not so fast.
That same Clearhead study revealed that while most online retailers have a personalization technology, few know how to make effective use of it. Of the 144 ecommerce executives interviewed for the study, only 17 percent had a plan/path to develop personalized experiences.
“Similar to testing, we find that lack of technology isn’t the problem when it comes to personalization. While 64 percent of retailers have the technology, they tend to lack the process and rigor needed to execute their personalization efforts,” Clearhead Co-Founder and Executive Chairman Sam Decker explains in the recently published Digital Optimization Benchmarking Study.
What makes personalization so hard to implement as a strategy? Let’s say you have the technology, you have loads of usable audience data to latch on to, and you’ve got ample time. After all of that, these are three big hurdles that marketers will have to overcome before personalization will become mainstream:
1. Before personalizing, how do you segment your audience?
One of the first questions a marketer needs to answer is how to slice the audience into groups that are worth personalizing for in the first place. There is no point in spending hours coming up with personalization strategies for audience segments that do not exist or are not real customers.
Even the task of identifying which data points to use can be daunting: What matters more -- that it’s nighttime somewhere or that it’s raining outside? Does gender matter or is age more important? The list goes on, and there isn’t enough time or site traffic in the world to test all the options.
Not to mention, this task becomes all the more difficult when these segments are constantly shifting with the seasons.
The quick solution: Work backwards. Think of the data point for which you’ve got a good personalization idea and that is relevant, and start there. For example, maybe your visitor a returning visitor or they entered your site thought a specific keyword search -- should the content you show them be generic? Half the challenge is “personalizing” based on a data point, so at least you’ll be halfway there.
The long-term solution: Keep testing alternate data points to use as the personalization criteria.
2. What does a personalized experience look like?
When Optimizely launched its personalization tool in mid-2015, it made it look so simple. CEO Dan Siroker showed how you could use a Skymosity integration to know what the weather was like where your visitor was located and customize a homepage banner based on that data point. So, when it was sunny for the visitor to your site, you could show a surfer with this season’s board shorts on and when it was rainy, you could show someone in the rain, wearing a poncho. Personalization accomplished! People in sunny places want board shorts and if it’s raining out, they will want to buy a poncho. Aren’t humans a simple bunch of apes? In practice, it’s not that simple (and Optimizely is well aware).
Other than showing return visitors product categories they already looked at instead of whatever is on sale that day, the strategy for what you do in response to any data point is highly unclear, as soon as you go beyond basics.
So what if you know that I listen to podcasts, work in Manhattan and just got married? Will any of that help you figure out how to best sell me something? Because I am a man, does it mean I like to see other men wearing clothes? Perhaps I’m shopping for my wife. How can you know? Are you going to rely on my returning to the website after showing interest in something, therefore only personalizing for a fraction of visitors?
Even if you find a general approach for a specific data point, are you sure that your changes are working well for the reason you think they are? Or is there another factor that you’re not taking into account?
With so many questions, it’s easy to see how the average marketer doesn't have the patience to deal with personalization, when there are more clear-cut approaches sitting in front of them.
The quick solution: Keep it simple. Men do tend to shop differently from women. Past behaviors may be indicative of future interests.
The long-term solution: Keep testing to find more effective experiences. Focus on where you get more bang for your buck, rather than attempting to create the most personalized experience possible.
3. How does personalization scale?
Let’s assume you’ve tackled all of the above and you’ve found an effective personalization strategy -- how do you apply that to all seasons, devices and products lines? Deciding how much of your conversion funnel has personalized content and how much is static will become a challenge rather quickly. Tracking what is working and what is not also becomes difficult. For example, say you’ve got one audience segment that is underperforming in conversion rates. Does it mean that your personalization strategy is off or that this segment simply is less apt to convert?
Making sweeping changes across a site can cause havoc to a personalization setup that is already in place, forcing you to start from scratch in some cases.
The quick solution: Make sure to not invest heavily in personalization where it risks being upended after a short while.
The long-term solution: Use personalization strategies as a way to learn about your visitors rather than purely a way to increase marketing return on investment. The value of insights tends to be evergreen, unlike quick conversion wins.
The secret to truly understanding and implementing effective personalization throughout marketing efforts lie in understanding user behavior and their profile. Once marketers fully understand customer preferences and the data points associated, they are able to better segment this information, creating the opportunity for authentic communication.