When it comes to user segmentation—the first of the four core pillars of personalization—you can’t have a one-to-one conversation with each customer without first breaking your audience down into neat, little segments.
In this segmentation process, though, there has to be intention. It would be peak inefficiency to look at your audience and go one-by-one tailoring a specific message for each individual. While a potentially great community or relationship builder, the lack of scale and automation would drive your team mad.
Instead, we take a look at similarities (and differences, as you’ll see) to build segments, which make a marketer’s life easier and a consumer’s experience more personalized. A true win-win situation.
What Segmentation and Cooking Have in Common
That said, sometimes we are prone to treating segmentation like a new recipe we’ve just found. You follow each step perfectly and use every suggested ingredient because the recipe has a proven track record of success without really questioning whether these elements are right for you or your dinner guests.
Segmentation, much like any recipe, is malleable and has to be contextualized to be most effective. Your business needs and those of your users must be taken into consideration when building out segments. Otherwise, one or both of you will be left with a sour taste in your mouth.
So, with that in mind, we’ve compiled a list of three pitfalls of segmentation that commonly arise from traditional methods and explain how you can effectively avoid them.
(Pro tip: Hate reading? Check out the video at the bottom of this post to learn more about these pitfalls and the strategies to avoid them.)
The 3 Pitfalls of Segmentation to Avoid
1. Correlation vs. Causation
You’re likely familiar with the difference between correlation and causation, but when looking at segmentation strategies, it might not be as obvious. In this situation, we are talking about understanding and contextualizing the outliers in customer data.
Let’s start with an example: Maybe you’re an individual who loves buying men’s clothes. Regardless of your gender expression, you are a loyal customer of your favorite brand and have been buying men’s clothes for years. As you might expect, you are receiving promotions for menswear. Nothing out of the ordinary here.
Put actions up against one another to see patterns.
But this last year, you purchased a dress for a special someone in your life who loves a classy sheath, dipping into women’s clothing from this brand for the first time. Now, with the most recent piece of data in mind, the brand sends you a stream of promotions for dresses and other women’s clothing items.
Clearly, you have entered a new segment based on the correlation between “bought a dress” and “buys dresses.”
Rather, segmenting based on causation seeks to understand the context and reasoning behind a purchase. In this way, the data can be used to effectively delineate messaging and products in the correct context for each individual.
A quick remedy for this segmentation pitfall could be something like a post-purchase pop-up or polling email that asks the customer about the reason for the purchase (either for yourself or as a gift).
In knowing the purpose of the purchase, the marketer can place that individual in the right segment, such as “primarily purchases men’s apparel, but receives recommendations for women’s gift items.”
2. Historical vs. Current Events
A key piece of segmentation is the historical data that marketers have in the user’s profile—what they like and what they have done. Sometimes, though, this can leave out a big piece of contextual information: what is happening in the world around the user!
This year, more so than most, has been a shining example of weighing historical data against current events. Consumer priorities have shifted. General dispositions have changed and will continue to change at a rapid pace. Marketers have to keep up, and part of keeping pace is incorporating current events into your consideration for segments.
The example below shows how Taco Bell is shifting to accommodate in-store dining closures in certain regions and the discomfort some diners feel in regions where restaurants have reopened. To help out, they’ve given users several choices to suit their preferences and contextualized this against historical data of the user’s favorite items.
Show customers that you’re paying attention by acknowledging their current situation.
This segmentation strategy continues as users make selections for their dining preferences (delivery vs. in-store) and subsequent messaging takes this into account.
3. Similarities vs. Differences
As mentioned above, the creation of a segment often hinges upon similarities—people who like hats, lovers of sushi, residents of Austin, TX, you name it.
But it’s also important to take into account the other side of the coin to avoid this common segmentation pitfall. After all, our differences are what make us unique, and understanding a user’s uniqueness makes personalization that much more effective.
This pitfall in particular touches on segments that are sliced too broadly, capturing a wider swathe of preferences without considering motivations. Segments are often made and whittled down using dichotomous actions (customer vs. non-customer; free vs. paid), but in doing so, intrinsic behaviors are brushed aside.
Consumers make actions based on factors, like how products make them feel or how they fulfill their needs. The most effective segments find a way to touch on these motivations.
Take meditation as an example. Some meditate to reduce anxiety. Others meditate as a form of mental fitness. While these might seem similar—and for some, there might be an overlap—these are entirely different motivations that warrant different approaches from a brand like Calm, which you can see in the messages below.
While the action might be the same, the motivation behind it could differ.
By recognizing motivation and subtle differences, such as these, messaging takes on greater relevance and resonance with users, leading to increased engagement and retention.
Using this tactic, Calm saw a 4x increase in revenue through testing and optimization of its onboarding sequence for new members.
Avoiding These Pitfalls of Segmentation
There’s no shame in seeing these pitfalls and admitting you fall victim to them. They’re subtle and easy to miss, but they can be avoided with a few safeguards.
When considering where to start, always remember to:
- Put the customer at the center of your strategy rather than brand, price or promotion.
- Emphasize context and personal values as the keys to one-to-one relationships.
- Take environment into account. It will help you understand customer needs and priorities.
For your data, it’s important to:
- Continually assess your requirements to identify what’s needed to drive personalized experiences.
- Refresh and confirm customer data regularly to verify its validity and relevance.
- Use progressive profiling to constantly learn more about your users.
And finally, it’s just good marketing to report the progress of your segmentation strategies (even weekly, if you mean business), so you can iterate where needed and optimize your results.
So there you have it! The most common segmentation pitfalls and a few simple steps you can take to stay vigilant and avoid them.
If you’re interested in diving deep into the topic of user segmentation, you should take a look at our recent Personalization Playbook that shows you how to build the perfect audience!
