Segmentation should be practical

A segment is a technical word to talk about customer groups. Cultural, societal, and economical factors are some of the data points that make them up. Creating customer segments is crucial when it comes to consumer understanding, and because of it, the key to generating sales. 

When conducting segment research, the aim is usually to get group percentages or fractions. These numbers tell you which part of the analyzed population belongs to a specific segment. But segments are not only a list of metrics meant to catalog people, segments are alive. This is the reason why, to represent real-world data, segments have to be created with archetypes in mind. Archetypes are categories created based on human behavior.

Archetypes, segments or personas?

Personas, archetypes, and segments are often used to refer to the same thing. Although sometimes used as synonyms, each of these words represent different concepts.

Personas and archetypes are used to talk about forms of segmentation. A persona is more of a design tool, whereas behavioral archetypes represent motivations, goals, and attitudes of the consumer. For example, think about creating the next Macbook. One of your personas might be a hip Gen Z’er who enjoys coworking spaces, and another might be a young designer. However, when we talk about archetypes, behavioral and quantitative insights are added into the mix. 

Shifting paradigms

Because people are at the core of business design, customer segmentation is needed to deliver meaningful solutions. Contrary to common beliefs, big chunks of data don’t equal better insights – here’s an example why.

A while back, a client of ours hired an external agency to conduct large quantitative surveys in order to get their hands on large-scale customer segments. Simultaneously, we conducted our own segmentation at a smaller scale. And the outcomes didn’t match. The customer segments, created by large surveys, didn’t match the behaviors we synthesized through interviews and digital experimentation, observing real in-market behavior.

When companies hire agencies with the sole purpose of delivering customer segmentation, they invest millions in surveys that collect information from thousands of people around the world – this still doesn’t guarantee reliable results. How is this possible?

Strategies backed up by hard data are a must. But it doesn’t matter how many people you get information from. In the end, it comes down to how you do it. If flawed, even the largest surveys will produce inconclusive results.

The flaws of traditional segmentation

It takes longer to create archetypes through traditional segmentation. This is not necessarily because the steps are wrong, it has more to do with the process itself: traditional segmentation is linear. 

Traditional research-based segmentation usually starts by interviewing a small group of people (gathering qualitative data), followed by conducting a very detailed quantitative survey afterwards. These long surveys are pushed to thousands of people across the world. After results come in, the final segments are created. 

The issue with traditional segmentation is that it claims to be behavioral because of some of the questions asked. But a long questionnaire ends up being counterproductive – asking behavioral questions doesn’t make your research behavior-based. 


Behavior = motivation + ability + triggers

Customer archetype segmentation

Think back to our previous example. We conducted customer research on a significantly smaller scale, and despite sharing common grounds when outlining the segments, there was a key differentiator: our process was a loop.

While our process also began with a series of interviews, we made sure we didn’t conduct them only once. Another highlight was that we didn’t wait to create segments until the research was over, we did it from the get-go. Iterations led to improved versions, and resulted in final segments based on real-world human characteristics called behavioral traits. 

Working with smaller groups of people allowed stepping away from multiple-choice answers, leading to deeper consumer insights. Meanwhile, several rounds of interviews increased the accuracy of the archetypes that represented each of our segments. By following a more behavioral-driven and iterative approach you can capture key learnings early on, and thus de-risk your entire segmentation approach.

The importance of iteration

Different approaches to segmentation don’t alter the steps to get from point A to point B. In reality, high-fidelity segmentation is the result of altering the order in which each step is implemented, and the number of repetitions you’re willing to pursue. 

Once your archetypes are in place, you should continue to implement resources to understand what triggers them and drives them to take action. Specifically, typing tools. A typing tool is a simplified set of questions that can automatically classify consumers into existing segments. Creating an initial typing tool early on in the segmentation process helps you finetune the segments and start identifying the demographic information that is required to calculate market sizing and solution potential. By continuously refining the typing tool alongside the segments you create a feedback loop that improves the segmentation while ensuring an accurate and usable typing tool. 

The feedback loop, the information gathered from iterations, and typing tools helps create a richer, final version of each segment, one that gets closer to what drives people in real life. 

Targeted digital validation

So far, the data that shaped your archetypes is a direct reflection of interviews, questionnaires, and the typing tool you’ve implemented – and the fidelity of your work will rest on iterations that validate these insights. 

Running digital experiments to (in)validate key assumptions and observing real world responses to different triggers helps to define archetype specific design criteria. This criteria will eventually inform your solutions.

This is where the typing tool comes in handy; by targeting segments on digital platforms (like social media), and getting them to participate in surveys using the typing tool, we collect more data through the social targeting to further improve and validate the segments.

What happens next?

As the number of people you research increases, so does the fidelity and depth of your insights: patterns and segments get closer to representing reality. Defining the groups of people you’ll be targeting is a solid first step towards imagining the products, services, and businesses your customers will need tomorrow, and for you to start creating them today.