---
title: "Savings & Waste"
canonical: "https://cloudmonitor.ai/docs/using-cloudmonitor/reports/savings-and-waste/"
description: "Every rightsizing and waste opportunity across the estate, ranked by what it saves, with a way to work through them."
---

Every waste and rightsizing opportunity across your estate, ranked by how much it saves. This is
where you go when you have time to spend well rather than a fire to put out.

## Four views of one list

**Savings & Waste** is everything. Three tabs slice it:

| Tab | What it finds |
|---|---|
| **Right-sizing** | Resources running larger than their workload needs — drop a size or enable autoscale |
| **Idle & Orphaned** | Idle services and orphaned resources still on the bill — deallocate or delete |
| **Storage Optimization** | Tiering, replication and snapshot hygiene |

Same underlying recommendations, filtered. Start on the combined list, sort by saving, and work down
— the ranking already tells you where the effort pays.

## The numbers

**Potential monthly saving** is what you'd save **If all open actions applied**. It's an upper bound,
not a forecast: some recommendations won't be appropriate for your workload, and that's what
dismissing them is for.

**Open opportunities** counts only items still open — postponed, dismissed and completed ones are
excluded, so the number reflects work actually outstanding.

**Resources affected** counts distinct resources, so one resource with three recommendations counts
once.

:::tip[These are estimates, not invoiced amounts]
A projected saving is modeled from usage patterns. It is not money that has moved, and the cost
basis chip is grayed out here because an estimate has no billed-versus-amortized version.

Treat the ranking as reliable and the absolute figures as indicative. If a number is going in a
board pack, say what it is.
:::

## Working through them

Each row has a **⋮** menu with the same triage actions used across CloudMonitor:

- **Comment/Send** — reassign to someone with a note. Opens a pre-filled email containing the item
  and a link back to it.
- **Postpone** — snooze for a day, three days, a week, two weeks or a month. Moves to Postponed.
- **Dismiss** — with a reason, minimum a few words. Moves to Dismissed.
- **Reopen** — brings a postponed or dismissed item back.
- **Open in Azure portal** — straight to the resource.

Anyone in the relevant cost group can triage. Everyone can open the resource in Azure.

**There is no "mark as done".** Once you act on a recommendation, it stops being detected and closes
itself on the next run. Dismissing is for "we've considered this and we're not doing it" — and the
reason is what stops the same argument being had again in three months.

## Why there's no date filter

Recommendations don't have a window. A finding is either current or it isn't. CloudMonitor hides the
Period chip rather than showing one that does nothing.

Provider and cost group both apply — they match the affected resource — so you can hand a team
exactly their list.

## Where cost optimization sits

If you're looking for the Well-Architected cost pillar, it's this screen.
[WAF Recommendations](/docs/using-cloudmonitor/reports/waf-recommendations/) covers the other four
pillars — reliability, security, operational excellence and performance — and links here for cost.