Concept 1 of 2

Segmenting customers and users

6 questions test this

A customer is whoever decides to acquire the product. A user is whoever has to live with it afterwards. In consumer products they are usually the same person. In almost everything else they are not, and a product owner who conflates them will keep being surprised by adoption numbers that do not match sales numbers.

Approaches to segmenting

By job to be done. Group people by the progress they are trying to make, not by who they are. Two people in different industries who both need to reconcile invoices before a month end are in the same segment for your purposes.

By behaviour in the product. Group by what people actually do, such as frequency, depth of use, or which path they take through a workflow. This is observable rather than reported, which is its main advantage.

By role in the buying decision. Economic buyer, technical evaluator, end user, and the person who has to approve the change. Each has a different question and each can stop the purchase.

By context of use. Same job, different environment. A field engineer with intermittent signal and an office user on a fixed desk need different things from the same feature.

Whichever approach is used, a segment is only useful if it changes a decision. If two segments would produce the same ordered backlog, they are one segment.

Handling conflicting needs

Conflict between segments is the normal case, not an exception, and the useful technique has three moves.

First, state the conflict in terms of the underlying need rather than the requested solution. Two requests that look incompatible often turn out to be one need with two guesses attached.

Second, check the conflict against the Product Goal. The goal already says who the product is currently for. If the goal cannot settle it, the goal is too vague, and that is the more valuable thing to have found out.

Third, decide and record the decision where the losing segment can see it. A visible ordered backlog does this by itself. The point is not that everyone agrees, it is that nobody has to guess.

Connecting developers to customers and users

The objectives ask for at least three approaches, and these are the ones that survive contact with a real schedule.

Developers attend the interviews. Not as note takers, as a second listener. The questions a developer asks are different from the questions a product owner asks, and cheaper to answer in the room than in a ticket.

Direct observation of real work. Watching someone use the product, or use the spreadsheet the product is supposed to replace, produces detail that no written requirement carries.

Developers at the sprint review. Stakeholders and users inspecting the increment with the people who built it removes one layer of translation.

Support and usage data made available, not summarised. Raw tickets and session recordings are more useful than a monthly digest, because the developer notices things the summariser filtered out.

The product owner is accountable for value, not for being the only channel to the market. A product owner who becomes the single point of contact with users has created a bottleneck and a game of telephone at the same time.

Common misconceptions

Customer and user are two words for the same person.

They coincide in consumer products and separate almost everywhere else. A procurement lead who signs the contract and a warehouse operative who uses the scanner want different things, and a backlog that serves only one of them fails in a predictable way.

Segmenting by company size or industry is enough.

Firmographic labels are convenient because the data already exists, not because they predict behaviour. Segments built around the problem people are trying to solve tend to hold together better, because two companies of the same size can have nothing in common.

Conflicting needs are resolved by building both.

Building both defers the decision and doubles the surface area. The technique that works is to make the conflict explicit, name which segment the Product Goal currently serves, and decide in the open.

6 questions test this concept

A product owner for a warehouse management system finds that the procurement director who signed the contract wants detailed reporting, while the warehouse operatives want fewer taps per scan. How should this be framed?

  • AThese are a customer and a user with different needs, and the product owner has to decide which the Product Goal currently serves and make that decision visible to both.
  • BThe procurement director's needs come first because they hold the budget.
  • CThe needs should be averaged so both groups get half of what they want.
  • DThe operatives are not stakeholders because they did not buy the product.
Check whether it stuck.

One per page, with a worked explanation.

Start the set
Related material
Book
Competing Against Luck, On grouping people by the job they are trying to get done.
Book
The Mom Test, On getting useful answers out of customer conversations.
Book
Interviewing Users, On running the interviews themselves.
Book
Crossing the Chasm, On choosing one segment and serving it completely.