Figure 12 Sept 2026From “False positives in churn models have a cost”

Customer churn intervention cost and decision matrix

Proactive retention outreach carries severe unit costs for false alarms. Sleeping dogs and stable accounts must be shielded from unneeded discounting.

Cite Embed

Figure 1Customer churn intervention cost and decision matrix

Classification state	Reality	Intervention	Cost profile	Customer response	Net outcome
True Positive (Saveable)	Real risk; solvable blocker.	Discovery and support fix.	High CSM labor hours.	Resolves blocker and renews.	Positive ROI: Contract saved.
True Positive (Lost Cause)	Irreversible risk (bankruptcy).	Discounts and escalations.	High labor and concessions.	Churns regardless of offer.	Negative ROI: Concessions lost.
False Positive (Stable)	Healthy account; seasonal dip.	Preemptive discount offer.	Unneeded margin loss.	Accepts unneeded discount.	Negative ROI: Margin destroyed.
False Positive (Sleeping Dog)	Inactive; inertia account.	Outreach citing low usage.	CSM outreach hours.	Cancels after wake-up call.	Severe Negative ROI: Caused churn.

Swipe or scroll horizontally if the table is wider than your screen.

Source: Author's commercial retention model grounded in Ascarza, Iyengar and Schleicher (2016). All figures and costs are synthetic decision benchmarks.

PNG4:3, 1600 × 1200, with the title, the source line and the attribution set in the image isoglu-2026-false-positives-in-churn-models-have-a-cost-fig01.png

Source and evidence

Each line names a claim related to this exhibit, its source, and how it was checked.

A AIS16-C1 The backfire, verbatim: encouraging customers to switch to cost-minimising plans can "increase rather than decrease customer churn"Ascarza, Iyengar & Schleicher. (2016)
A AIS16-C2 The proposed mechanisms are the authors' own: the campaign increased churn "by lowering customers" inertia to switch plans and by increasing the salience of past-usage patternsAscarza, Iyengar & Schleicher. (2016)

Grades: A, verified against the printed page of the primary source · B, primary source, text layer only · C, authoritative secondary · D, reported.

From the essay

Revenue operations & AI

False positives in churn models have a cost

Machine learning churn models prioritize recall, but false alarms trigger unneeded discounts, wasted customer success hours, and wake-up churn effects.

2 Sept 2026

Read the essay

This exhibit belongs to an essay: the numbers live here, the argument lives there.

Cite or embed

APA 7

Isoglu, S. (2026). Customer churn intervention cost and decision matrix [Figure]. isoglu.com. https://isoglu.com/insights/exhibits/isoglu-2026-false-positives-in-churn-models-have-a-cost-fig01/

BibTeX

@misc{isoglu-2026-false-positives-in-churn-models-have-a-cost-fig01,
  author = {Isoglu, Sinan},
  title = {{Customer churn intervention cost and decision matrix}},
  year = {2026},
  howpublished = {Figure 1, isoglu.com},
  url = {https://isoglu.com/insights/exhibits/isoglu-2026-false-positives-in-churn-models-have-a-cost-fig01/},
  note = {From the essay {False positives in churn models have a cost}}
}

Embed this exhibit

Copy this HTML to use the PNG exhibit on your site with its source line and a link back to the exhibit page.

<figure>
  <a href="https://isoglu.com/insights/exhibits/isoglu-2026-false-positives-in-churn-models-have-a-cost-fig01/">
    <img src="https://isoglu.com/social/isoglu-2026-false-positives-in-churn-models-have-a-cost-fig01.png" alt="Figure 1: Customer churn intervention cost and decision matrix" width="3200" height="2400">
  </a>
  <figcaption>Figure 1: Customer churn intervention cost and decision matrix. Source: Author&#39;s commercial retention model grounded in Ascarza, Iyengar and Schleicher (2016). All figures and costs are synthetic decision benchmarks. <a href="https://isoglu.com/insights/exhibits/isoglu-2026-false-positives-in-churn-models-have-a-cost-fig01/">View exhibit</a>.</figcaption>
</figure>