> ## Documentation Index
> Fetch the complete documentation index at: https://docs.assetinfinity.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Reorder recommendations

> What to buy, how much, and why — with the demand history the recommendation was built from.

What is running out, ranked by how much it matters.

<img src="https://mintcdn.com/assetinfinity/95UxrhuglB-erOSy/images/inventory/reorder.png?fit=max&auto=format&n=95UxrhuglB-erOSy&q=85&s=66091d7d6b957d050fbb4b9a904d490a" alt="The reorder recommendations list" width="2880" height="2550" data-path="images/inventory/reorder.png" />

## What a recommendation is

Not just "stock is low". A recommendation weighs:

| Input                        | Why                                                                |
| ---------------------------- | ------------------------------------------------------------------ |
| **Current stock**, per store | One empty store is a shortage even if the total looks fine         |
| **Reorder point**            | The level you decided is too low                                   |
| **Consumption rate**         | What you actually get through, from the ledger                     |
| **Planned demand**           | What upcoming [maintenance plans](/maintenance/plans) will consume |
| **Lead time**                | Six weeks to arrive means ordering six weeks earlier               |
| **Open reservations**        | Stock already promised to jobs                                     |

Planned demand is the part people miss. A part with three on the shelf and a plan that consumes four
next month is short *now*, even though nothing has gone below its reorder point yet.

## Reading the reason

Every recommendation shows the argument behind it — the rate it is consuming at, how that rate was
measured, what is planned, and when stock runs out at that rate. The reason travels with the
recommendation, so somebody approving a purchase can see why rather than trusting a number.

Demand rates carry a quality mark. A part with two years of steady history gives a rate you can plan
on; one with three movements in eighteen months gives a guess, and says so.

## Acting on it

* **Raise a purchase request** for the recommended quantity.
* **Transfer** from another store that has spare, rather than buying.
* **Substitute** — the [alternatives](/inventory/part-detail) on the part may already be in stock.
* **Dismiss**, with a reason, where the recommendation is wrong. It stays visible rather than
  vanishing.

## Seasonality

Where consumption has a seasonal shape — a filter that gets changed far more often in the monsoon —
the rate reflects it rather than averaging it flat across the year.

## Keeping it honest

Recommendations are only as good as what is recorded. Parts issued against jobs build the
consumption history; parts taken off a shelf without being issued do not, and the forecast drifts
low. Issuing against the job is what keeps this screen worth reading.


## Related topics

- [Inside a part](/inventory/part-detail.md)
- [Inventory settings](/config/inventory.md)
- [The part catalogue](/inventory/parts.md)
- [Finding your way around](/get-started/navigating.md)
- [Reliability](/assets/reliability.md)