Key takeaways
- A weekly report exists to make decisions. Lead with spend, revenue, MER, cost per purchase and ROAS, then explain the change, then list actions.
- Compare full weeks, pulled on the same weekday, and against a four-week average as well as last week, so noise and attribution delay don't mislead you.
- Break ROAS into CPM, link CTR, conversion rate and order value to see which lever moved.
- Check Meta-reported revenue against store revenue every week. MER is the referee when platforms disagree.
- End every report with three to five decisions for the coming week, each tied to a number in the report.
A good weekly Meta ads report fits on one page and leads to decisions. It shows spend, revenue, MER, cost per purchase and ROAS for the full week against last week and the four-week average, explains what moved using funnel and creative metrics, and ends with three to five actions for next week. Pull it on the same day each week so attribution delay affects every week equally.
Why report weekly?
Daily numbers are noisy. Purchases arrive late, weekends behave differently from weekdays and a single large order can swing a day's ROAS. Monthly reports are too slow: by the time you see a problem, you've paid for four weeks of it.
A week is long enough to smooth daily swings and short enough to act on. It also matches Meta's own rhythm: the learning phase is measured over seven days, and the default click-through attribution window is seven days.
What should a weekly Meta ads report include?
Five sections, in this order. Most readers stop after the first, so make it count.
- Headline numbers. Spend, revenue, MER, purchases, cost per purchase, ROAS and new-customer share.
- Why it moved. The funnel metrics that explain the change: CPM, link CTR, conversion rate, average order value.
- Campaigns and creatives. What drove spend and results, which ads are winning, which are fatiguing.
- Tests. What was tested, the verdicts, and what's graduating to scaling.
- Decisions. Three to five specific actions for next week.
The template
Here's a template table for the headline section. The numbers are a hypothetical example.
| Metric | This week | Last week | Change | 4-week average | Note |
|---|---|---|---|---|---|
| Amount spent | $7,000 | $6,000 | +17% | $6,200 | Budget raised on scaling campaign |
| Store revenue (all channels) | $26,500 | $24,000 | +10% | $24,300 | From Shopify |
| MER | 3.79 | 4.00 | −5% | 3.92 | Store revenue ÷ total ad spend |
| Meta-reported purchases | 175 | 160 | +9% | 158 | 7-day click, 1-day engage, 1-day view |
| Cost per purchase | $40.00 | $37.50 | +7% | $39.24 | Target $42 |
| Meta-reported ROAS | 2.38 | 2.51 | −5% | 2.42 | Break-even 1.90 |
| Average order value | $95 | $94 | +1% | $95 | Purchase value ÷ purchases |
| New-customer share | 68% | 71% | −3 pts | 70% | From store data |
If you sell through other channels, MER uses total ad spend across all of them. Here we've assumed Meta is the only paid channel, so total spend equals Meta spend.
The diagnostic section
| Metric | This week | Last week | Change | What it tells you |
|---|---|---|---|---|
| CPM | $14.00 | $13.00 | +8% | Auction cost |
| Link CTR | 1.30% | 1.35% | −4% | Whether ads earn the click |
| Cost per link click | $1.08 | $0.96 | +12% | CPM ÷ CTR |
| Landing page view rate | 82% | 83% | −1 pt | Page loading for clicks |
| Conversion rate (per landing page view) | 3.2% | 3.1% | +3% | Site and offer |
| Frequency (prospecting, 7 days) | 1.9 | 1.7 | +12% | Audience saturation |
The creative section
List your top five ads by spend, with cost per purchase, ROAS, hook rate and frequency. Flag any ad whose cost per purchase has risen for two weeks running while its frequency climbs. Those are your fatigue candidates. Our guide to ad fatigue explains the signals.
How do you pull the numbers correctly?
Consistency matters more than precision. A few rules keep weeks comparable:
- Use the same date range every time, for example Monday to Sunday, and pull it on the same weekday, such as Tuesday.
- Remember attribution delay. Ads Manager credits purchases to the day of the click or view, so the last days of the week keep filling in. Pulling every report on Tuesday means each week has had the same time to settle. Our guide to attribution settings explains why.
- State the attribution setting in the report and keep it the same across ad sets you compare.
- Use the same time zone as your ad account and store, so days line up.
- Save a reporting preset in Ads Manager so the columns don't drift. See Meta Ads Manager metrics for presets worth building.
How do you read week-over-week changes?
Start by asking whether the change is real, then which lever caused it.
Is the change real or noise?
- Compare with the four-week average, not only last week. A week that's down 10% on an unusually strong week may be perfectly normal.
- Check the sample size. A cost per purchase change on 20 purchases is far less reliable than one on 200.
- Look for calendar effects. Paydays, holidays, festive sales, a long weekend or a product launch can explain a swing on their own.
- Check for tracking changes. A sudden drop across every campaign, while store orders stayed steady, points at the pixel or Conversions API, not the ads.
Which lever moved?
Break ROAS into its parts. For any period:
Revenue per 1,000 impressions = link CTR × conversion rate × average order value × 1,000
ROAS = revenue per 1,000 impressions ÷ CPM
Here conversion rate is purchases per link click. For example, if ROAS fell 5% while CPM rose 8% and CTR, conversion rate and order value held roughly steady, the drop is mostly the auction, not your ads or your site. If CPM held and link CTR fell, look at creative. If CTR held and conversion rate fell, look at the site, the offer or stock.
| Pattern | Likely cause | Decision |
|---|---|---|
| CPM up across the account, other metrics steady | Seasonal competition | Hold budgets, protect target CPA, don't panic-edit ads |
| CTR down on top ads, frequency up | Creative fatigue | Launch new variants, retire tired ads |
| CTR steady, conversion rate down | Site, price, stock or offer change | Check the product page and checkout, test the funnel yourself |
| Order value down | Mix shift to cheaper products, offer ended | Revisit bundles and thresholds |
| Meta ROAS up, MER flat | Meta claiming sales that would have happened anyway | Check new-customer share and retargeting spend |
| Meta ROAS and MER both up | Likely genuine improvement | Consider scaling winners in steps |
| Everything down sharply, store orders steady | Tracking problem | Fix tracking before touching campaigns |
Use MER as the referee
MER, total store revenue divided by total ad spend, ignores attribution entirely. If Meta's reported results and MER move in the same direction, trust the change. If they diverge, find out why before acting. The MER calculator and ROAS calculator help with the arithmetic.
What decisions should come out of the report?
A report without decisions is just a record. End with three to five actions, each linked to a number above. Typical decisions:
- Scale: raise budget on ad sets beating target CPA by a clear margin, in steps such as +20 to 25%.
- Hold: keep budgets steady where CPA is close to target or the ad set is still learning.
- Cut: reduce or pause where CPA has been above target for two weeks and isn't improving.
- Refresh: launch new creatives for ads showing fatigue.
- Test: start the next creative or offer test, with a clear success threshold. Our creative testing framework has the structure.
- Fix: tracking, page speed, stock or checkout issues spotted in the funnel.
Write each one as an instruction with a number: "Raise Scale campaign budget from $500 to $600 a day, because cost per purchase was $34 against a $42 target for two weeks."
Decisions to avoid
- Changing everything at once. If you refresh creative, change budgets and edit audiences in the same week, next week's report can't tell you what worked.
- Reacting to one bad day. Daily drops are mostly noise.
- Cutting on Meta ROAS alone when MER and store revenue disagree.
- Scaling losers back to life. More budget rarely fixes a weak ad.
What does a good weekly summary look like?
Above the tables, write two or three sentences that a busy founder can read in ten seconds. Using the hypothetical week above:
Spend rose 17% to $7,000 as we scaled the main campaign. Store revenue rose 10% and cost per purchase held at $40, under the $42 target. MER slipped from 4.00 to 3.79, mostly because CPMs rose 8%. Next week: hold budgets, launch four new creatives for the two ads showing fatigue, and fix the slow product page.
A good summary does three things. It states the result in plain numbers, it names the main cause, and it says what happens next. It doesn't list every metric, and it doesn't hide bad news in the tables.
How long should the report take?
Once your column presets and template are set up, the report should take about 30 to 60 minutes a week. If it takes longer, you're probably pulling numbers by hand from too many places. Keep a spreadsheet with the template tables, paste the week's exports into it, and let formulas fill in the change and four-week average columns. Save the time for the analysis and the decisions, which are the parts that actually move results.
What to do when the report shows a bad week
Don't rush. Check tracking first, then the calendar, then the funnel. If the cause is clear, fix that one thing. If it isn't, hold budgets for another week rather than making several changes at once, and look at the trend over two weeks before acting.
Who should read the report, and how?
Tailor the depth, not the numbers.
| Reader | What they need |
|---|---|
| Founder or owner | Headline table, two-line summary, decisions |
| Marketing team | Full report, including creative and tests |
| Finance | Spend, store revenue, MER, cost per new customer |
| Agency or freelancer | Full report plus notes on what they changed |
Keep a running log of weekly reports in one place. Over a few months, it becomes the best guide you have to your account's seasonality, typical CPMs and creative lifespan.
Fly ROAS keeps a plain-English Flight Log of every test verdict, stop and scale it makes in your account, which fills the tests and decisions sections of this report for you if you use its Autopilot.
A weekly reporting checklist
- Full week pulled, same weekday as last time, same attribution setting
- Headline table with this week, last week and the four-week average
- Store revenue and MER checked against Meta-reported revenue
- ROAS broken into CPM, CTR, conversion rate and order value
- Top five ads reviewed for fatigue
- Test verdicts recorded and winners graduated
- Three to five decisions written, each tied to a number
Summing up
A weekly Meta ads report is a decision tool, not a data dump. Show the headline numbers against last week and the four-week average, use MER to keep Meta honest, break ROAS into its levers to find the cause, and finish with a short list of specific actions. Pull it the same way every week, and over time it becomes the clearest record of what really works in your account.
Frequently asked questions
What should be in a weekly Facebook ads report?
Lead with spend, store revenue, MER, purchases, cost per purchase, ROAS and new-customer share. Then add the funnel metrics that explain changes, such as CPM, link CTR, conversion rate and order value, a review of top ads, test results and three to five decisions for next week.
How do I compare Meta ads performance week over week?
Pull full weeks on the same weekday so attribution delay affects each one equally. Compare with last week and with the four-week average, check sample sizes, look for calendar effects, then break ROAS into CPM, CTR, conversion rate and order value to see which lever moved.
Why do last week's Meta results keep changing?
Ads Manager credits a purchase to the day of the click or view, not the day of the sale. With a seven-day click window, purchases keep arriving for days after the week ends. Pulling every report on the same weekday keeps weeks comparable.
Should I report Meta ROAS or MER?
Both. Meta ROAS shows which campaigns and ads perform best inside Meta. MER, total store revenue divided by total ad spend, ignores attribution and shows whether the business as a whole is getting more efficient. When they disagree, investigate before you act.
How many decisions should a weekly ads report include?
Usually three to five, each tied to a specific number in the report. More than that and you change so much at once that next week's report can't show which change worked. Fewer, well-reasoned decisions give cleaner learning.
