“A lot of products are one-size-fits-all. And it doesn’t work super well for anybody. And nothing is done about it.”Indi Young, Author of Practical Empathy and Mental Models

Chances are that you’ve been building your product to address the needs of your best fit customers, or personas prioritized by your team.

But, just as there are variations in your user base, there are also variations in the profiles within the groups you’re targeting—accountants from Ohio won’t all share the exact same needs or processes.

At maturity, one-size-fits-all isn’t optimal.

There could be hundreds of sub-segments of users using your product. Some users within these segments may be completely satisfied with your product, while others might be looking for a replacement.

By creating experiences that better match the needs of these sub-segments, you’ll be able to:

  • reduce churn and dissatisfaction in neglected segments;
  • increase usage and revenue across your user base; and
  • define more tailored acquisition strategies to grow the most promising sub-segments.

But just as there are different ways to cut a cake, there are also different ways to resegment a customer base.

The 4 Types of Customer Segments

Good segments are homogeneous, mutually exclusive, have predictable behaviors, and can be reached through marketing and sales efforts.

The earlier in the customer journey that you are able to segment users, the better you will be able to adapt the experience to the specific needs of each segments.

So, what you’re looking for is the earliest signal–leading indicator that allows you to predict behaviors across segments.

Good segmentation criteria might be company size for a chat product, or user roles for a sales tool, or number of subscribers for Mailchimp.

There are four main ways to resegment users:

  1. With implicit data: Implicit data is information that’s inferred from other available data. As an example, sign-ups originating from organic channels tend to have higher purchase intent and convert better. Applying this theory to acquisition channel data could help with segmentation.
  2. Explicit data: Explicit data is information captured through a form, a survey, or a setup process. Users understand that they are being asked questions, to which they give answers.
  3. By user or buyer persona: Personas can be very powerful when they are based on provable theories (e.g. with real causation).
  4. With a behavioral model: Behavioral segmentation leverages transaction and engagement data to create profiles.

To determine which segmentation criterion your business should use, you should evaluate the correlation between different events and user attributes.

You’re looking for criteria that are most likely to correlate with purchases, engagement, or long-term retention.

Finding the Best Criteria to Help Resegment Your Customer Base

You can test different recipes by combining criteria (e.g. persona AND revenue). It’s often a good idea to focus on information that’s unlikely to drastically change in the near-term, for example:

  • Employee headcount
  • Revenue
  • Vertical
  • Business model
  • Location(s)
  • Tech stack
  • Web traffic
  • Use case
  • Acquisition channel
  • Landing page value proposition
  • Persona
  • Role
  • Job title, etc

What criteria seem most predictive?

For each of the most predictive criteria, explore the resulting segments: How different are they from one another?

You’re looking for, at most, five to seven groupings. Your segments should be mutually exclusive, cover most—if not all—of your customers, reveal underlying needs, and ideally be reliable long-term.

How to Learn From Your Segmentation

No matter which recipe you choose, it’s important to stick to a single segmentation model.

Select the most promising segmentation model. Break down your metrics by segments: are behaviors noticeably different?

If you don’t understand what drives these differences, then consider exploring segments one by one through customer interviews.

If you’re not already collecting all the information you need, then look for ways to collect it through the sign-up process, the onboarding flow, or data enrichment using external services like Clearbit Enrichment.

Once you are capturing the right information, you’ll be able to experiment with workflows to influence the behavior of the users in each of your segments. Resegment until you find the right targeting.

– –

This post in an excerpt from Solving Product. If you enjoyed the content, you'll love the new book. You can download the first 3 chapters here →.

Categories: Customer Research Technique