← All Labs

Conversion Propensity

Audience
01

Customer question

Which customers are most likely to buy?

02

What this solution does

Instead of treating every customer the same, this ranks them by how likely they are to buy soon, so marketing spend goes toward the people most worth reaching.

Specifically, it answers: Given behavioural signals about a customer, how likely are they to convert in a given window, so marketing can prioritise effort and spend accordingly?

03

How it works

Behavioural, engagement, and historical purchase variables are fed into a classification model that outputs a conversion probability per customer, which can then be segmented into propensity deciles for targeting.

04

Data requirements

  • Previous purchase history
  • Website/app visit and engagement data
  • Recency and frequency of interactions
  • Product/category view history
  • Marketing engagement history
  • Category interest signals
05

Discovery questions

What conversion event is being predicted, and over what time window?

Propensity models must be trained against a clearly defined outcome and horizon, or the score won't mean what stakeholders assume it means.

How much historical conversion data exists to train against?

Classification models need enough labelled positive and negative examples to learn a reliable pattern.

How will the propensity score actually be used in campaigns?

Determines whether the output needs to be a simple decile split, a continuous score, or feed directly into bidding.

06

Build an experiment

07

Generate synthetic data

08

Data readiness

Generate a dataset in the previous step to run readiness checks.
09

Run model / simulation

Generate a dataset first to run the model.
10

Results

Run the model in the previous step to see results here.
11–12

Business ⟷ Technical interpretation

The top propensity decile has an average conversion probability of topDecileProbability%, compared to bottomDecileProbability% in the bottom decile — a strong signal for prioritising media spend toward higher-propensity segments.

Shown as a template — the highlighted tokens are filled in once the model has been run.

13

What would you do next?

  • Pilot a campaign targeting only top-decile customers and compare efficiency against a broader send
  • Monitor propensity model performance over time and retrain periodically
14

Solution & vendor landscape

Implementation approaches

Platform-native predictive audience toolsInternal data science / ML pipelineCDP with built-in predictive scoringSpecialist marketing analytics vendor
See the full Vendor Landscape →

Related concepts