Columns, verdicts and Pro mode on Ad performance
This page explains the What's working table on Performance > Ad performance. It covers every column, the six verdict badges and the Pro detail level. For the page itself, see Ad performance.
Problem and intended user
This page is for the media buyer (the person who runs your paid ads) and the creative strategist (the person who plans and makes ad creatives). A long list of ads with many numbers is hard to read. It is hard to tell which ads deserve more budget, which cost you margin, and when creatives wear out. A wrong call keeps money on weak ads or retires a creative too early.
Current workflow
You compare ads by hand. You export results from Meta Ads Manager or Google Ads to a spreadsheet and add formulas for return, cost per acquisition and share of spend. You track creative age and launch dates in a second sheet, or ask an agency for a creative report.
Prerequisites
- Wait for the first ad sync to finish.
- Your order and customer data in Commander. Predicted revenue return and True profit need it. Without it, both columns show a dash, and you hover the dash to read why. A verdict then falls back to ROAS.
- To use your own definition of a winner in Pro mode, set Winning criteria. Without it, Commander compares each ad with your other ads.
How Commander works
Fixed rules produce every column, verdict and score here. Nothing needs an approval, and nothing changes your ad accounts.
Compute the columns
Select Metrics to show or hide columns. Name is always on. The first column shows a concept, hook, ad or other group, depending on the view. Commander reads spend, revenue, transactions, video plays and reach from your synced ad data. It adds your order data for Predicted revenue return, True profit, New customers and nCAC.
Decide each verdict
Commander judges every verdict with fixed rules, and the page only shows the result. A verdict compares a row with your account average return. The row uses Predicted revenue return when Commander has order data for the row. Otherwise it uses ROAS.
- Per ad. An ad is judged on its own return. In By ad, and for the ads inside a group, each ad gets its own verdict. A group row is judged on the return of all its ads together, and it also needs enough ads. A group that holds a single ad, such as a Performance Max asset group, gets the same verdict as that ad.
- Always against the whole account. The comparison uses your whole account, never only the ads on screen. A spend, status, format, source or search filter does not change a verdict.
- Too little to call. A row with too few purchases in the dates you picked stays at Not enough data whatever its return says. The Transactions column shows the purchases Commander counts. A group with only two ads, or with too few purchases across its ads, stays at Not enough data too.
- Scale and Toxic need a real share of your spend. A small ad can be Strong or Leaking. It does not become Scale or Toxic.
- A loss stops Scale. A row whose True profit is negative never gets Scale. It gets Strong instead.
- Needs classification. The Unclassified group holds ads Commander could not place on the current view. In a creative view, those are ads Commander has not analysed yet. It gets no verdict. It shows Needs classification.
- No verdict on the other fixed groups. In a creative view, the Performance Max, No benefit claim and Unplaced groups get no verdict. The Verdict cell shows a dash.
- No verdict on a link group. In By campaign and By ad set, the No campaign linked and No ad set linked groups hold ads Commander could not tie to a campaign or ad set. They get no verdict. The Verdict cell shows a dash.
The verdict does not use your winning criteria. That setting defines what counts as a winner elsewhere. See Winning criteria.
Switch to Pro mode
Use the Simple and Pro buttons at the top of the page to switch between the two. The headline numbers stay. The What's working table gives way to a Creative fatigue section. It reads from the same dates and platform as the rest of the page.
Exact output
Columns
| Column | What it shows |
|---|---|
| Status | What Meta or Google reports for the ad now. On an ad only. |
| Campaign | The campaign the ad runs in. On an ad only. |
| Ad set / Ad group | The Meta ad set or Google ad group. On an ad only. Off by default. |
| Creative type | Static, Video, Carousel, Text or Shopping & Performance Max. On an ad only. |
| Source | Commander when Commander launched the ad from a creative it generated, Other otherwise. On an ad only. Off by default. |
| Spend | Total spent in the period. |
| Revenue | Revenue the ad platform attributes to the row in the period. |
| ROAS | Platform-reported return: Revenue divided by Spend. |
| Predicted revenue return | Commander's estimate of your customers' lifetime revenue divided by ad spend. It measures revenue, not profit. Needs your order and customer data. |
| True profit | Profit on the orders Commander matched to the row, minus the part of its spend those orders cover. Needs your order and customer data. |
| CPA | Cost per acquisition: spend divided by transactions. |
| CPC | Cost per click: spend divided by clicks. |
| CPM | Cost per thousand impressions: spend divided by impressions, times 1,000. Off by default. |
| New customers | First orders Commander matched to the row in the period. A Google match can count as part of an order. Needs your order and customer data. |
| nCAC | New-customer acquisition cost: spend divided by New customers. Needs at least one whole first order. Needs your order and customer data. |
| Hook rate | 3-second video plays divided by impressions. Meta video ads only. Needs 1,000 impressions on days with video data. Off by default. |
| Hold rate | ThruPlays divided by 3-second plays: of the people the video hooked, the share that kept watching. Meta video ads only. Off by default. |
| ThruPlay rate | ThruPlays divided by impressions. Meta video ads only. Off by default. |
| Cost per ThruPlay | Spend on the video ads divided by ThruPlays. Meta video ads only, with video data on every day of the range. Off by default. |
| Reach | Daily reach of the row's Meta ads, added up over your dates. Meta only. Off by default. |
| Frequency | Impressions divided by Reach. Meta only. Off by default. |
| CTR | Click-through rate: clicks divided by impressions. |
| Impressions | Times the ads were shown. Off by default. |
| Clicks | Clicks on the ads. Off by default. |
| Transactions | Purchases the platform attributes in the period. |
| Spend share | The row's share of the spend on individual ads. On by default. Google campaign-level spend with no ad rows is not in the total. |
| Ads | How many creatives the row rolls up. |
| Launched | The day Commander first saw the ad. On an ad only. Off by default. |
| Trend | A small line of the row's trailing 7-day return over the 30 days that end on your end date, with the change as a percentage. In the By ad view, the Trend column shows a dash. |
| Verdict | How the row compares with your account average return. |
INFO
Predicted revenue return is a forecast Commander fixes when it matches an order to an ad. It does not follow the dates you pick.
WARNING
True profit counts only matched orders, so it is smaller than the full picture. If your profit figure already subtracts marketing cost, True profit counts that cost twice. For store-wide profit, use Gross profit after ad spend on the page.
How the calculated columns work
Select the information icon beside a column name to see its formula. The Metrics menu shows the same formulas.
- ROAS is Revenue divided by Spend. The revenue is the ad platform's own figure.
- Predicted revenue return is the revenue Commander predicts the customers of the row will spend over their lifetime, divided by Spend. Commander bases it on customer lifetime value (CLV). It counts each customer once, and it covers matched orders only.
- True profit is the profit of the matched orders, minus Spend times the matched spend share. The matched spend share is the revenue of the matched orders divided by the revenue the platform reports for the row. Commander charges only that part of the spend, so the profit and the cost describe the same orders.
- New customers counts the matched orders that were a customer's first order. nCAC is the Spend of the ads behind that count divided by the count, the same basis as CPA. A campaign or ad set leaves out the spend of an ad whose first orders Commander cannot tell apart.
- Hook rate, Hold rate and ThruPlay rate add up the plays and impressions of the row's Meta video ads first, then divide. A group's rate is never an average of its ads' rates. Cost per ThruPlay uses the spend of those video ads only, and shows a dash when a day in the range has no video data.
- Frequency is the impressions of the row's Meta ads divided by their Reach.
Matched orders are the orders Commander could tie to a specific ad. The four numbers at the top of the page cover your whole store, so Gross profit after ad spend and True profit differ by design.
Verdicts
Select the information icon beside Verdict to read this legend in the table.
Scale, Strong, Leaking and Toxic also tint the left edge of the row. On par and Not enough data add no tint. The label always shows, so you never need the colour alone.
| Verdict | Meaning |
|---|---|
| Scale | Far above your account average and holding, on a meaningful share of your spend. Put more budget here. A row whose Trend is falling does not get Scale, and neither does a row with a negative True profit. |
| Strong | Above your account average. |
| On par | Close to your account average, so it is not clearly better or worse. |
| Leaking | Below your account average. It costs you margin. |
| Toxic | Far below average on meaningful spend. It is the biggest drain in the account. |
| Not enough data | Too few purchases, or too few ads in a group, to call yet. |
Pro mode sections
The Creative Health Score card rates your creative mix as Strong, Healthy, Watch or Needs Attention. It shows three bars: Monthly Churn, Age Mix and Hit Rate. The overall score averages Monthly Churn and Hit Rate. Age Mix is not included, because it rises when more ads stop early.
A recommendation under the score offers See more and See full analysis. Below it, the three metric cards and a fourth Creative capacity card each open one section:
| Section | What it shows |
|---|---|
| Monthly Creative Churn by Launch Cohort | How spend from each monthly launch cohort decays over time. Oldest cohorts sit at the bottom. |
| Creative Age Mix — Daily Spend Composition | What share of daily spend comes from ads of each age. The darkest shade is the newest creatives. |
| Testing Quantity & Hit Rate | Weekly creative output by type and win status, with a hit rate trend line and a list of tested ads. |
| Creative capacity | How fast creatives stop delivering, how long winners last, and how many new creatives to launch to keep up. |
The tested ads list labels each ad Winner, Potential winner or Not a winner. When you set Winning criteria, the hit rate uses them to decide what a winner is. Otherwise it compares each ad with your other ads. A note appears if the criteria changed while the launches were judged.
The Media Audit also reads the winner verdicts that Commander stores for your ads. Its winners follow your winning criteria, or the default criteria until you save yours.
Ads that ran
The Ads that ran list appears under Monthly Creative Churn by Launch Cohort and under Creative Age Mix — Daily Spend Composition, when you open either section. It has the columns Ad, Platform, Start, End, Days lived, Status (Active or Inactive), Spend, Revenue and ROAS. Money shows in your ad account's own currency. Use Previous and Next to page through the list. Select an ad name to open its Ad detail page.
If no ad spent in the period, the list reads No ads spent in this period. If it fails to load, it reads Couldn't load the ads that ran in this period. with a Retry button.
What to do with it
| What you see | Decision or action | Owner |
|---|---|---|
| Scale | Put more budget here. Change the budget in Meta Ads Manager or Google Ads, outside Commander. | media buyer |
| Strong or On par | Keep it running and watch its Trend. | media buyer |
| Not enough data | Wait for more purchases before you decide. | media buyer |
| Leaking or Toxic | Review Toxic first, the biggest drain. Lower, pause or rework its ads in your ad platform, outside Commander. | media buyer |
| Needs classification | Switch to By ad, or wait until Commander has analysed more of your creatives. | media buyer |
| Creative Health Score at Watch or Needs Attention, or a Creative capacity result | Plan new creatives to replace those that stop delivering. | creative strategist |
| A Winner in the tested ads list | Reuse its idea in the next brief. | creative strategist |
Expected value and measurement
These views can help you act sooner on ads that lose margin and find creatives that last. They show a signal. They do not guarantee that a change will pay off.
To check a decision, open Performance > Ad performance. Compare the row's ROAS or Predicted revenue return and its Verdict over two equal date ranges, before and after your change. In Pro mode, compare Hit Rate and the Creative Health Score the same way.
You cannot isolate one change when budgets, offers or seasons moved at the same time. A verdict compares a row with your whole account, so changes in other ads can move it. A row can move between On par and Strong without any change to its own ads.
Limitations and exceptions
A verdict is a prompt. Treat a verdict as a prompt to look closer, not as a final call.
Estimates. Predicted revenue return is a forecast. True profit covers matched orders only, and it can count marketing cost twice.
Order data missing. Predicted revenue return, True profit, New customers and nCAC show a dash when Commander has no order and customer data for the row. Hover the dash to read why. Verdicts then use ROAS.
New customers. The count comes from the orders Commander matched, not from Meta or Google. A dash means Commander cannot tell first orders apart for your store or for that ad. nCAC shows a dash below one whole first order or without spend.
Video and reach columns. Hook rate, Hold rate, ThruPlay rate and Cost per ThruPlay cover Meta video ads only. Static ads, carousels and Google ads show a dash, and so does a row below the 1,000-impression floor. Cost per ThruPlay also shows a dash when a day in the range has no video data. Reach and Frequency cover Meta ads only. Commander adds up daily reach, while Meta counts each person once, so Reach reads higher and Frequency lower than Meta Ads Manager on a range of more than one day.
Thin data. The whole Creative fatigue section reads Not enough data yet to score creative fatigue. Choose a longer date range or wait for more ad history. The Creative Health Score band can read Not enough data. A Based on N of M signals line can appear when only some of the signals have data.
Load errors. Couldn't load creative fatigue data. and Couldn't load creative capacity data. each come with a Retry button.
Demo Mode. The data is sample data.
A card in Pro mode can show its own message instead of a score:
| Card message | When it appears | What to do |
|---|---|---|
| Not enough launches yet | Monthly Churn needs 2 complete months with 3 or more ads each. | Wait for more launches, or choose a longer date range. |
| No ads ran in this period | Age Mix found no ad with spend in the period. | Choose a different date range. |
| No tested launches yet | Hit Rate found launches that are still below the test threshold. | Wait until your launches finish testing. |
| Can't measure this range | The range holds too much history to load at once. | Narrow the date range. |