Effective User Segmentation

Product-led growth (PLG) is often thought of in terms of growth hacks, minor optimizations, self-serve flow, and connecting product-qualified activities (PQAs) to sales. There’s nothing wrong with these tactics and in many cases are important aspects of moving KPIs. At the same time, these tactics are not the core of a product-led growth approach. These are symptoms of the PLG approach, not the fundamentals. You will find many growth leaders brainstorming product hacks or optimizations to move KPIs. A ton of time is spent on experimentation. I have been a culprit in using that approach for a long time. The challenge is this creates a short-term approach and not a long-term sustainable growth motion. You quickly run out of low-hanging fruits and the journey gets harder to find impactful optimizations.

The first principle for any successful PLG approach is to understand your users and customers well. You want to go beyond defining ICPs and know where in the product journey are each of these users. The starting point of a PLG strategy should be building the product and the GTM motion in such a way that it understands the users in an automated manner and presents targeted touch points to graduate users across the journey (Journey being acquisition, activation, engagement, monetization, etc). I translate that first principle to an effective User Segmentation.

Okay, so how do you go about segmenting your users and accounts?

A huge prerequisite to building effective segmentation is to have an account 360 view. This involves maintaining good hygiene of data across a user journey including:

User behavior data: Understand users’ behavior in the product. Essentially boils down to event tracking. Typical events include, activation point tracking, signup tracking, upgrade tracking, trial tracking etc. Amplitude, Mixpanel, HotJar are some common tools to enable user behavior tracking.

Marketing automation and sales data: You want to capture marketing automation data including interaction with email and ad campaigns. Then add the sales data. This typically comes from Salesforce and includes fields like LedGen score, opportunity stage, etc.

Account and billing data: Now, connect the account data and the billing state. These are typically saved in the company’s database or through payment services like Stripe. You want to know what plan tier a customer is on, their ARR, etc.

Company profile and Use cases: You want to know the company size, industry, user’s job role, and their use cases. These are typically captured through an onboarding form or through Clearbit/Zoominfo APIs.

That’s a lot of data. Many companies end up putting these together in a data warehouse (Snowflake, Bigquery, Redshift, etc) and running queries on them. Keeping these data points connected and clean is whole another motion. That’s probably a topic for another day.

While not in perfect shape, many companies do have this kind of data now. So, the next step for them is to create account segments. A typical behavior is to ask your data team to run queries and build some interesting user segments. There are some major inefficiencies in that approach:

Growth leaders don’t always know what segment they need. So, everyone starts to guess and this becomes a months-long exercise.

It takes a long time for the Data team to perfect the query and get you the segments

These segments quickly become static: Typically the query outcomes are downloaded into a CSV and then uploaded into other tools (e.g. Marketo) or share with growth engineers to run some experiments.

Often the outcome of the queries lives on a spreadsheet or a presentation to the product team. There’s little action to improve the user experience

If any of that sounds familiar, you are not alone. I am curious to learn how you are tackling some of these pitfalls.

If done right (which I think some B2C companies do e.g. Uber, Instagram, etc), this segments-based approach to create user touch points across the user journey can do wonders in driving product-led growth. Uber has more than 1500 segments and they are constantly targeting these segments both inside and outside their app. Those many segments sound like a managing hell but an automated way to create effective segments and connect across user touch points can build sustainable growth.