AR23, strategic and AI consultancy
PLEN

Market communication

Analytics audit

Data has been coming in for years, and nobody in the company can say what users actually use. We put the measurement in order and show how the product is used.

“We have been collecting data for years. Which product features are needed, and which merely exist?”

Note for decision-makers

Result
A map of actual product usage and a measurement plan that makes it possible to assess the next change.
Makes sense when
The company is planning a product change: a rebuild, slimming down, a new feature. And wants to know afterwards whether it worked.
Does not make sense when
The question is “why do users do this”. That is answered by talking to people, not by data.
The first question we will ask
What decision is this data meant to support? Data collected without a question rarely answers one.

What you get

  • Measurement put in order in the tools the company already has (GA4, PostHog, Hotjar).
  • A feature usage map, with reach and depth counted separately.
  • A breakdown by account type and entry route, because the average user does not exist.
  • Transactional features separated from informational ones. Each is measured differently.
  • A measurement plan before the next change: what we collect, for how long and what level we consider a success.

Two principles

Niche does not mean dead. A feature is assessed by the reach and depth of its use. Example: a feature used by half a percent of users, but in sessions lasting several dozen minutes, is both niche and essential. Removing it on the basis of its share of page views alone is a common mistake when slimming down a product. Only the people whose working tool has been taken away notice it.

Measure before, not after. To know whether a change worked, data from before the change has to be collected before it happens. Afterwards, all that is left is the after measurement, and a story takes the place of a comparison.

How the work proceeds

  1. 01

    We check that events are collected completely and without double counting. Without that, analysis makes no sense.

  2. 02

    We build the feature usage map.

  3. 03

    We divide users into groups that use the product differently.

  4. 04

    We set the measurement plan for the next change.

What you can count on

A picture of what is happening in the product and on what scale. The reasons behind user behaviour are explored in interviews, described under the pre-implementation review.

What the price depends on

The state of current measurement (sometimes most of the work is repair), the number of tools to put in order, the size of the product and the number of features. Whether only a picture of the current state is needed, or a measurement plan as well.

Let's talk about your case.

The shortest route is one sentence about what you are trying to decide. We will get back to you on whether and how we can help.

Tell us what change to the product the company is planning.

Write to us →