A swing electorate, under a policy reversal.
How fast approval flips from with-you to against.
reports
Two kinds of reading. Stress tests are real runs on real populations, published in full, no login. Field notes are longer-form thinking on how populations move, and why most tools miss it.
real runs · no login · no sales call
The hard reports. Each one points the engine at a real population, runs the pipeline, and publishes what fell out. These are the lead magnet and the proof of horizontality at the same time: a new domain published here is the breadth argument, demonstrated rather than asserted.
9,030 typed defection edges; 63% land where no win-back can follow.
1,599 typed defection edges; 13% land where no win-back can follow.
Resale anxiety at 0.90 defection-proximity, and the exits already named.
Public signal flagged a ~14% at-risk cohort, most of it reversible.
How fast approval flips from with-you to against.
Where latent attrition becomes a wave.
longer-form thinking
The thought-leadership stream. Essays on populations, tipping points, and the difference between counting complaints and modeling a crowd. This is where the category gets defined in words, and it is deliberately styled as writing, not data.
A single brand is too sparse to hold structure. Build the friction map from the category, then score the brand against it.
Every churn tool counts dead bodies. Here is why that is the wrong unit of measure.
Source, destination, trigger. A defection is a sentence, and the parts are the part worth keeping.
Some friction never ages. Some expires in months. Treating both as current is the central error of complaint-scraping.
Linear alerts on a non-linear process. The math of being surprised.
A forecast tells you which customers resemble the ones who already left. It cannot tell you who leaves when you change the conditions.
What Foundation got right about modeling people, and where the honesty has to come in.
Leaving is a crossing, not an event. Mid-crossing there is a window where a small intervention does the work of a large one. Most retention budgets miss it.
new stress tests and notes land here first. the whole argument in one place is the churn guide.
Bring us a population