Skip to main content

Overview

Where did your Meta budget go, and which slice of the audience paid it back?

90 second readAppears on: /marketing/meta/analytics

Meta spends your money on your behalf every day, and decides for itself who sees the ads. This page is where you check its work.

Every number here is Meta's own reporting, pulled from your ad account. The sales it credits to an ad are Meta's claim, not your Shopify till — use the Marketing page when you need the version your bank agrees with.

The cards​

Settle two things before you open any breakdown: whether the account paid back, and what one order cost you.

ROASROASRevenue earned for every ad dollar spent.4 or more is healthy · CPACPAWhat one order costs you to win through ads. · SpendSpendWatch it to control budget and back the winners. · CPMCPMWhat it costs to reach a thousand people. · CTR (Link)CTR (Link)How enticing your ad is to the people who see it.1.5% or more is healthy · CPC (Link)CPC (Link)The cost of each site visit from Meta. · FrequencyFrequencyA read on ad fatigue in your audience.Under 2 is healthy · ReachReachYour true audience size; compare with frequency for saturation. · Link clicksLink clicksTraffic Meta is sending; judge it by the sales it converts.

They are the total for whatever the filters at the top currently allow — every campaign and every country unless you narrowed it.

Read them in three moves​

  1. Did it pay back? ROAS, then CPA. ROAS tells you the ratio, CPA tells you the price of one order.
  2. What did it cost to be seen and clicked? Spend, then CPM, then CPC (Link).
  3. Is the audience wearing out? Frequency against Reach, then CTR (Link).

Three pairs worth checking​

Frequency + CTR (Link). Frequency rising while CTR (Link) falls is the same people being shown the same thing until they stop noticing. New audience or new creative — a bid change will not fix it.

Reach + Spend. Spend up with Reach flat means the extra money bought more showings to the people you already had, not more people.

ROAS + CPA. ROAS holding while CPA climbs means bigger orders are quietly paying for a more expensive customer. That works until order size stops rising.

The tables​

Somewhere in your account is a country, an age band or a placement that the rest of your budget is paying for. This is where it shows itself.

Country, Age, Gender and Placements cut the same nine measures a different way. Only the meaning of a row changes.

Every tab opens sorted by Spend, highest first, and every measure has a % △ beside it comparing to the previous period.

When Meta cannot say which country, age or gender a result belongs to, it lands in a row called unknown. That row is real spend — do not skip it.

Country​

Meta picks which countries to serve, not you. This is where you find out whether it picked well.

One row = one country Meta served your ads in.

How to scan it

  1. Sort by Spend, highest first.
  2. Look across at ROAS, then at the total row.
  3. The row that matters: a country in your top three for spend with ROAS below the total. It is being carried by the others.

Age​

Your cheapest customers are usually in a band you are underfunding. Moving budget between age bands costs you nothing but a click.

One row = one age band.

How to scan it

  1. Sort by CPA, lowest first.
  2. Look across at Spend.
  3. The row that matters: the cheapest band getting one of the smallest budgets. That is a shift you can make today without new creative.

Watch the % △ on Frequency here. An age band that Meta keeps returning to will fatigue before the others do.

Gender​

If your creative speaks to one audience while your money goes to the other, this is the first place it shows.

Usually two rows and an unknown. Small tables mislead, so read shares rather than sorting.

How to scan it

  1. Compare Spend between the rows — that is what Meta chose, not what you asked for.
  2. Read CTR (Link) across the same rows.
  3. The row that matters: one taking most of the budget on a weaker CTR (Link). Your creative is speaking to one audience while your money goes to the other.

Placements​

The cheapest fix on this page. Turning off a bad placement needs no new creative, no new budget and no new campaign.

One row = one placement, with a Publisher platform column naming the platform it ran on.

How to scan it

  1. Sort by Spend, highest first.
  2. Read CTR (Link) and CPA together on the top rows.
  3. The row that matters: a placement taking real budget with a CTR (Link) far below its neighbours. Creative built for a feed rarely works in a story slot.

Filters​

Get these wrong and every number above is answering a question you did not ask.

The bar at the top is Campaign: All and Ad Set: All.

The ad-set list only fills in when exactly one campaign is picked. Choose more than one and opening it says "Select a single campaign in the campaign filter to see the ad set filter options".

Changing the campaign clears the ad-set choice, because ad sets belong to a campaign.

These filters follow you. They stay on across Top Ads, Campaigns and Ad Formats, and survive a page reload. A filter set ten minutes ago on another Meta page is still narrowing this one.

An account selector appears at the right only when more than one Meta ad account is connected. Switching account reloads the page and clears the campaign and ad-set choice.

The date range at the top of the app applies. The country, customer type and source filters do not — Meta pages read only your dates and your shop. Use the Country tab for a country view.

Each table has a chart/table switch, an Export button and a Search box. In chart view you get a dimension picker, a metric picker of up to three measures, and a Compare tick box.

Watch out for​

Reach and Frequency do not add up. They count unique people, so a person in two rows is counted in both. Adding the column gives you a number that means nothing.

The total row is Meta's total, not the sum of the rows. It is calculated for the whole selection in one go, so rows Meta could not attribute never appear above it. Rows failing to add up to the total is normal here.

Read CPM as a direction of travel. Judge it by where it moves across the days in your range, not by the exact figure on one card.

Sales here are Meta's attribution. They count a purchase Meta believes it influenced within its own window. Your Shopify Total Sales on the Marketing page counts orders that actually happened. Neither is wrong and they will not agree.

Do this first​

Open the Placements tab, sort by Spend, and read CTR (Link) down the top five rows.

Placement is the one slice on this page you can change in an afternoon without touching creative or budget.

See yoursYour Meta spend, ROAS and CPA split by country, age, gender and placement.

Open Overview →