Comparison
Cause-specific hazard vs cumulative incidence for churn
Two correct ways to measure one exit type, answering two different questions
A side-by-side comparison of the two legitimate ways to measure one churn exit when other exits compete with it. The cause-specific hazard tells product and research teams what drives cancellation among accounts still on the books. Cumulative incidence, modelled with Fine and Gray's subdistribution hazard, tells finance how many accounts will actually cancel. Use it to choose the right curve before a board review. The split follows Austin, Lee and Fine's guidance on competing risks.
Where this appears
What's in this comparison
Cause-specific hazard
- Answers why accounts cancel
- Rate of cancelling among accounts that have not yet exited
- Suits product and research teams
- Use it when you are explaining a cause
Cumulative incidence (Fine-Gray)
- Answers how many accounts will cancel
- Absolute risk of this exit, given every other exit that can happen
- Suits finance, forecasting and save programmes
- Use it when you are forecasting a number
Use this image
Free to use in your docs, decks and posts under CC BY 4.0, as long as you credit Koji with a link to this page. To embed it, paste this into your page:
html
<a href="https://www.koji.so/visuals/cause-specific-hazard-vs-cumulative-incidence-churn"><img src="https://www.koji.so/visuals/cause-specific-hazard-vs-cumulative-incidence-churn/cause-specific-hazard-vs-cumulative-incidence-churn.png" alt="Comparison of two churn measures: cause-specific hazard explains why accounts cancel, cumulative incidence (Fine-Gray) forecasts how many will." width="1200" height="668" /></a>
<p>Source: <a href="https://www.koji.so/visuals/cause-specific-hazard-vs-cumulative-incidence-churn">Cause-specific hazard vs cumulative incidence for churn</a>, Koji</p>
