In-market timing

What it represents

Scoring answers who. Timing answers when. For each account it predicts which of the coming twelve months that account is most likely to enter a buying cycle, and surfaces the standout months as an in-market window.

It is a separate model type with its own configuration and its own runs, not a modifier on a score.

Timing is in beta, and the label in the product says so. It has been validated on synthetic seasonal data so far, not yet on live customer history. Treat the ordering it gives you as a planning input, not as a settled forecast.

How it’s calculated

Timing fits one survival model across your whole book of accounts. Each account’s history is expanded into a month-by-month timeline, and the model learns a monthly hazard: the likelihood an account enters market in a given month, given it has not already. Those monthly figures compose into a survival curve, and the months whose hazard stands out above the account’s own average become its peak months. Accounts that never had the target event carry through as open cases rather than as negatives.

Features for any month are drawn strictly from events before that month, so the model is never reading forward.

Confidence comes from how much signal an account has, not from routing to a different model. Accounts with rich history (roughly six or more events and at least two completed cycles) get high confidence. Three or more events gets medium. Sparse accounts are still scored by the same model, at low confidence, and their far-horizon predictions are faded rather than presented at full strength. Sparse accounts are covered because the model generalized from the whole book, not by matching them to named neighbors.

Timing refreshes on a dispatched run. There is no per-request timing call.

What it means for you

The peak month writes back to your CRM as a real date field, so filters like this month or next 30 days work natively in views your team already uses.

Worked example: across 4,000 accounts, 260 come back high confidence with a peak month inside the next 60 days. That is a workable planning list. Given the beta label, sanity-check a sample against what your team already knows before you build a quarter around it. Sample data cannot be written back to · writeback needs a connected CRM.