Skip to content
2 minute readNewAgent-ready

Prices that learn how your buyers respond

When you report whether a quoted price was accepted or rejected, Last Price now learns how sensitive your buyers are to price, per product and for the whole workspace, and the default optimizer can price from it instead of a generic guess.

Settings, Learned pricing, with the switch on and the estimates learned for all products and for each product

The default optimizer prices from how sensitive demand is to price. Until now it used the metadata.elasticity you sent, or a fixed assumption when you sent none, and most callers sent none.

  • Every night, quotes from the last 90 days that carry a reported outcome are used to estimate that sensitivity, for each product and pooled across your workspace, with a sample size and a 95% interval
  • An estimate is only used when it can be trusted: enough accepted and rejected quotes, prices that actually varied, and an interval clearly below zero. Otherwise the reason is shown, so you know what data is still needed
  • Settings, Learned pricing turns it on and lists every estimate in plain terms ("a 10% higher price wins about 17% fewer sales"), or why one is not used yet. It is off until you turn it on
  • Agents and scripts use the same switch and estimates through GET /api/learned-elasticity and PUT /api/learned-elasticity with {"enabled": true}
  • Learned prices move in steps: never far past the prices the estimate was learned from, so a slope measured over small price changes is not stretched to a large one
  • A metadata.elasticity you send still wins, and the explanation says when a price came from a learned estimate and how many quotes it rests on