Use Case

Prevent churn before it happens

Ask Dan which high-value customers are likely to leave. It finds the prediction that answers, checks it is ready to use and builds the audience, so your team acts before they go.

Predict

Churn risk and predicted value, for every customer.

Target

The high-value customers most likely to leave.

Act

A journey that reaches them before they go.

Dan at work

One question. The audience that answers it.

Which high-value customers are likely to leave? Dan finds the prediction that answers, checks its training quality before using it, and builds the audience on its scores. You create it with one approval.

Approval: Create this segment of 3,380 customers?

  • Readsyour predictions, their training quality and your definitions of value and churn.
  • Proposesthe prediction that answers your question, and the audience built on its scores.
  • Waits for your yesbefore a segment is created or a prediction starts training.

Every change Dan prepares waits for your review before it runs.

Meet Dan
In this use case

The problem

Churn shows upafter the fact

Customers rarely say they are leaving. They order less, then stop. By the time a report shows it, they have been gone for months, and a generic reactivation campaign reaches everyone at once.

A churn prediction changes the order of things: each customer has a churn risk and a predicted value, so the team knows who to reach first, and who is worth the effort.

What Dan prepares

The prediction, the audience,the next step

You ask in plain words. Dan reads your customer context and the predictions of DinMo Intelligence, then prepares:

01

Prediction

Dan picks the prediction that answers: here, LTV & Churn risk, which scores each customer’s churn risk and value over six months. It checks the training quality before using it: Excellent, 74.41% accuracy, in the example.

02

Audience

High-value customers to win back, built on the prediction’s scores: 3,380 customers in the example, 1.82% of all customers. The segment updates itself every day.

03

Journey

Once the audience exists, Dan can prepare the win-back journey that reaches it: the email for those who agreed to it, a reminder, a holdout.

04

Measure

The holdout receives nothing, so what the journey changes is measured against customers who were not contacted.

What you approve

Your question.Your decision.

Every change Dan prepares waits for your review before it runs. The segment exists only after you confirm “Create this segment of 3,380 customers?”, and a new prediction starts training only with your yes.

01

Dan proposes

The prediction that answers, and the audience built on it.

02

DinMo verifies

The prediction’s training quality, before its scores are used.

03

You approve

The segment, with one confirmation.

04

Every action is logged

Each action keeps who approved it and when.

Modules

The modulesbehind it

01

Intelligence

The churn risk and the predicted value of every customer, trained on your data.

Explore Intelligence

02

Audiences

The audience of customers at risk, refreshed every day.

Explore Audiences

03

Journeys

The journey that reaches them before they go.

Explore Journeys

04

Context Manager

Your definitions of value and churn, the same for every answer.

Explore Context Manager

FAQ

Questions,answered

Where does the churn risk come from?

From DinMo Intelligence: a prediction trained on your own data, in your warehouse, which scores the churn risk and the value over six months of every customer. Dan reads it; it does not make a score up.

What if the prediction is not good enough?

Dan reads its training quality before using it and tells you. If no prediction answers the question, it can propose one, and the training waits for your approval.

Does the audience update itself?

Yes. Once created, the segment updates every day: customers who come to match it enter, those who order again leave.

More use cases

Your next use case.With your data.

Let’s discuss your data, your tools and the use cases you want to build.