A frequent topic of confusion is Amazon failing to place Purchase Orders in line with their forecast, prompting the outcry of "Where are my POs?"
To better understand the likely timing of POs, you need to transform the Amazon forecast of end market consumer demand into something that will better represent the demand on the factory. This is a non-trivial calculation and is an excuse to re-use and mildly revise a response that I made to a LinkedIn Amazon Vendor Central post
Amazon's forecasts are forecasts of Amazon's end-market sales from a manufacturer perspective. The MktFc is the expected "P50" period unit sales that Amazon expects to have a 50% chance of being above or below. Similarly for the P80, Amazon expects only a 20% chance of market demand exceeding that value in the period. Until/unless you have a better forecasting method, to translate the chosen forecast into a meaningful estimate of potential incremental demand on a factory, on each ASIN you need to net off Amazon's excess inventory, Accepted but as yet unreceived POs, add in an expected Free Replacements%, your Sourcing % of Manufacturer sales, the spread the how Amazon typical orders across your Vendor Codes, Vendor Code-specific in-bound lead times (esp. for DI) and any adjustments you foresee for future pricing changes and/or promotions ...
I don't recommend attempting to guess the timing of Amazon POs, just make sure you have the right capability to supply, all assuming you want the business at the current price.
The historical part of this calculation is a tedious but automate-able calculation that Merchant AI does for 10s of thousands of ASINs per month. You are then left to add the future looking perspective e.g. promos that Amazon may know nothing about (yet).