What's the most effective way to use a behavioral segmentation (based on user product engagement) with a more psychographic segmentation (that looks at user needs, attitudes) and a more demographic segmentation to drive product growth and user engagement?

I love this question because it’s a challenge I’ve certainly faced in conducting segmentation and persona work. The reality is that different stakeholders need different things out of segmentations to make them actionable for their type of work and ideally you want different user models to be simple enough to ladder up to something the entire company can easily understand and get behind.

First, let’s talk about the benefits of each type of segmentation you mentioned. Then we’ll talk about how to make them connect.

Behavioral segmentation. Behavioral segmentation has three killer advantages: 

1) It’s highly actionable within your product and customer marketing. 

2) It’s based on attributes that you already collect and maintain on a regular basis. 

3) It allows you to model how you think about your users based on objective data and intent signals rather than surface level patterns and characteristics. 

• The downside is, you often won’t be able to identify or target users based on this set of criteria until later in the funnel.

Psychographic segmentation. Psychographic segmentation can be really helpful in a few ways: 

1) Identifying deeply with your audiences’ mindsets enables you to craft strong positioning, messaging, and onboarding that meets them where they are.

2) It adds important motivational and context clues that enrich personas that design can use to make more supportive design choices for your target audience.

3) It can help you size out what portion of your TAM is in a more “Serviceable Addressable” range, since many times, not all psychographic groups are equally ready to enter a consideration and purchase cycle with a solution like yours.

• The downside is, it’s extremely difficult to measure psychographic attributes. That said, you may be able to work with data partners to find correlations with other behavioral and market attributes.

Demographic segmentation. Depending on your business, demographic cuts might be incredibly predictive and helpful in identifying your target customers, or it might only be tangentially useful as a blunt targeting mechanism higher in the funnel. For example, if you’re a product designed for a particular life stage, gender, or geographic region, you’ll likely find a lot of success focusing here. If you have a more general use product, you may find that you want to limit the degree of demographic segmentation you rely on. Advantages to demographic segmentation include:

1) Much easier to apply to top-of-funnel and paid marketing targeting.

2) Easy for cross-functional stakeholders to understand and consistently use as cuts.

• The downside is, it can limit your reach and you don’t always get access to this information once people have signed up to your services. That can pose problems for continuing to action on this type of segmentation in your customer marketing, unless you explicitly ask. Some demographic details are considered PII and focusing too much on demographics can introduce biases your company may not intend.

How to connect different methodologies into one simple-enough user model

Now, my personal ideal is to be able to benefit from the best of all these approaches in a way that connects to a central understanding of your users that can be used company-wide. Here’s how I think about approaching this.

If your product is a tech product and you’ve been able to use cluster analysis to back into the most predictive attributes to help you bucket your users into groups that require different treatment, then you should lean on your behavioral segmentation as a central framework. 

If you conduct a separate psychographic market study, if you do enough due diligence, you might be able to connect your company’s data sets to the softer and more psychographic attributes you are measuring in your surveys. If so, you can bring all these things together more closely if you find any correlations.

As mentioned above, in some cases, demographic attributes may be the most predictive and cleanest for bucketing users based on the use cases you serve. If that’s the case, start with that demographic data at the center of your model and then analyze whether any behavioral attributes map more significantly to key groups. If not, you may want to use demographic segmentation for most of your marketing and just rely on behavioral data to help you with managing triggers and timing in your customer lifecycle marketing. 

There are times when it’s just not going to make sense to bring all of these methodologies together, sometimes they just don’t “fit” or have enough correlation to mean more when put together. In these cases, I recommend the following:

• Refer to your behavioral attributes as customer behavioral signals, and use it primarily for product and marketing lifecycle targeting and timing. 

• Refer to your psychographic patterned segments as a personas, and use them primarily when crafting messaging or to share with product and design as a potential design target.

• Refer to your demographic cuts as segments, and use them primarily for marketing targeting, especially on platforms where you have little other data to go off of. You may be able to bring in elements of the psychographic and behavioral trends into your segments to make them richer if there is significant enough correlation to do so. 

Embarking on segmentation projects can be daunting, but it helps to really clarify what your ultimate business goals are, who your internal audience is and how durable this segmentation needs to be. That can help you right-size your investment of time and strategic thought.