# Savings and recommendations

Read what is priced, and know on which cost base.

## The steps

1. Open **Savings**, under **Optimization**, and read **Total Annual Savings**.
2. Look at the calculation base, at the bottom of the page: your costs, or reference costs.
3. Open the **All Opportunities** tab, and expand a row to see its proof: current sizing and target, observed peaks, days powered off.
4. Open **Recommendations**, in the same group, and sort by urgency: that is the list to work through.

## Two pages, two uses

| Page | It answers | What it contains |
|---|---|---|
| **Savings** | How much can we recover? | Priced savings, in four families: reclaim, idle VMs, resizing, conversion |
| **Recommendations** | What to do, and in what order? | Capacity, cost, performance, security and resilience together, with an action plan |

## Three ratings to read an opportunity

- **Effort**: the workload to carry it out.
- **Time to Realize**: from under a month to over six.
- **Confidence**: how solid the proof is. Right-sizing based on observed peaks is more solid than a simple average.

Start with what combines large gain, low effort and high confidence: often storage reclaim and powered-off machines. Treat conversion as a project.

> An amount is there to compare opportunities with each other: it is not an invoice. Until your company has set its costs, it rests on reference values, and the page says so.

## Follow the result

A recommendation has no "done" box. It disappears by itself at the collection that shows the problem fixed. The **Annual Savings Trend** curve says whether waste is going down.

### Savings change from one day to the next

They are recalculated on your latest data: a machine powered off or deleted between two collections moves them.

### A machine recommended for resizing has peaks

Expand the row: the target takes the observed peaks into account, with a margin set in the **Configure Opportunities** tab.
