liveBook · Manning

Sleeping dog

As explained in Causal AI

A sleeping dog is a person who would experience the favorable outcome without a promotion or treatment but the unfavorable outcome if it were applied.

A sleeping dog is someone who achieves the favorable outcome without an intervention but experiences the unfavorable outcome when the intervention is applied. In churn prevention, the subscriber would stay without a promotion but leave after receiving one. section 8.3.3

Counterfactual subscriber segments distinguish who should receive an intervention:

  • Persuadables respond favorably to a promotion and can be retained. section 8.3.3
  • Sleeping dogs respond negatively, so intervention can create the unwanted outcome. section 8.3.3
  • Sure things stay without treatment, while lost causes churn regardless of treatment. section 8.3.3

A retention email that backfires

A subscriber who would have remained loyal may receive a marketing email, decide the company sends too much spam, and cancel. The email acts as the intervention that wakes the sleeping dog. section 8.3.3

Figure 8.5 In attempts to influence behavior, we break down the target population into these four segments. Given limited resources, we want to target our influence efforts on the persuadables and avoid the others, especially the sleeping dogs.
figure

Use the segment diagram to compare sleeping dogs with persuadables, sure things, and lost causes when allocating limited promotion resources. section 8.3.3

How Causal AI covers it

How deep it goes
Explains the segment through binary counterfactual outcomes and treatment contrasts, without specifying a fitting algorithm.
Examples
Uses promotions and marketing emails that cause otherwise loyal subscribers to cancel.
What it assumes
Assumes binary treatment and outcome reasoning while introducing the four customer-response segments.
What you can do after
Apply uplift modeling to target persuadables and avoid spending promotion costs on sleeping dogs.
How it connects
Connects sleeping dogs to uplift modeling, counterfactuals, persuadables, sure things, lost causes, promotions, and churn.

This explainer was generated by Manning's knowledge graph from the text of Causal AI. The section links point to the parts of the book it draws on.