# How to read Meta ads breakdowns

> Open Meta Ads Manager breakdowns by time, delivery and action, then read them without cutting the wrong placement or age group. Steps and a rupee example.
- **Author**: Marxx
- **Published**: 2026-11-20
- **Category**: Meta
- **URL**: https://marxx.ai/library-guides/how-to-read-meta-ads-breakdowns

---

In Ads Manager, select a campaign, ad set or ad, click **Breakdown**, and pick **By time**, **By delivery** or **By action**. Read the total first, then use the parts to ask questions, not to cut spend.

Each breakdown splits one total into parts, such as age or placement.

## How to open a breakdown

1. Go to Ads Manager and select the campaign, ad set or ad.
2. Click **Breakdown**.
3. Choose a group and then one breakdown inside it (checked 2026-10-20, Meta's help page).

| Group | What it splits by | Use it to ask |
|---|---|---|
| By time | Day, week, two weeks, month | Is the cost drifting up? |
| By delivery | Age, gender, country, platform, placement, device | Who and where did it reach? |
| By action | Conversion device, video view type | What did people do after the ad? |

Meta says you cannot combine two breakdowns from the same group.

## Know the limits of the numbers

- For offsite conversions, Meta supports these By delivery breakdowns: age, gender, age and gender, country, impression device, platform, platform and device, and placement (checked 2026-10-20).
- Meta says platform, platform and device, and placement values are estimated.
- Meta says it may use statistical modelling in breakdowns when conversion data is missing or partial.
- Conversions that Meta cannot place in a group appear as "Uncategorized".

So a small gap between two rows is not a finding. Look for big, repeated gaps.

## A worked example

Say you spend INR 50,000 on one campaign with [Advantage+ placements](/glossary/advantage-plus-placements). You open **By delivery**, then placement.

| Placement | Spend | Purchases | Cost per purchase |
|---|---|---|---|
| Reels | INR 20,000 | 40 | INR 500 |
| Feed | INR 22,000 | 55 | INR 400 |
| Stories | INR 5,000 | 4 | INR 1,250 |
| Other | INR 3,000 | 6 | INR 500 |

The rows add up: INR 20,000 + INR 22,000 + INR 5,000 + INR 3,000 is INR 50,000, and 40 + 55 + 4 + 6 is 105 purchases. The campaign cost per purchase is INR 50,000 / 105 = about INR 476.

Stories looks dear at INR 1,250. But it has only 4 purchases, and the system may have moved money there on purpose. This is the [breakdown effect](/glossary/breakdown-effect). If you want to know, run one test: copy the ad set, turn Stories off, and compare total cost per purchase after seven days. If the total does not improve, put Stories back.

## What to look for in each group

- **Time.** A cost that climbs week after week often means a tired ad.
- **Age and gender.** Use it to guide new creative, not to cut a group on a small sample.
- **Placement and platform.** Use it to decide which ratios to make. A weak Stories result may mean your ad does not fit a vertical screen.
- **Country or region.** Useful for budget when you target several states.
- **Device.** A big gap can point to a slow mobile page.

## Mistakes to avoid

- Turning off a placement because of its average cost
- Judging a row with a handful of purchases
- Comparing a breakdown total with a different date range
- Treating estimated values as exact

For a full account check, see [how to audit a Meta ad account](/guides/how-to-audit-a-meta-ad-account). To put breakdowns into a routine, see [how to build a weekly ad report](/guides/how-to-build-a-weekly-ad-report).

We make Marxx. It reads your connected Meta ad account in full and sends alerts and digests. See [creative intelligence](/features/creative-intelligence).

## FAQ

### Why do my breakdown rows not match my campaign total?

Meta says breakdown values can be estimated, and conversions it cannot place in a group show as "Uncategorized". Compare rows with each other and the total with your own sales data.

### Can I see more than one breakdown at a time?

Yes, from different groups, such as day and placement. You cannot combine two breakdowns from the same group.

### Should I cut my worst placement?

Not from a breakdown alone. Test it in a copied ad set and judge the total cost per result.

---
- [More Meta articles](https://marxx.ai/library-guides/category/meta)
- [All articles](https://marxx.ai/library-guides)