A useful Amazon Ads case study explains what changed, compares like-for-like results and separates attributed revenue from profit. A ROAS screenshot alone cannot tell a reader whether the campaign improved, whether the reporting window changed or whether a promotion drove the result.

This page provides a reporting template. It does not present an Arctavia customer success story or claim a measured improvement.

Start with the decision

Write the question before selecting a chart: “Did this keyword change improve efficiency while retaining sales?” is more useful than “Can we find our best ROAS week?” Name one primary outcome and the trade-off you need to observe.

Keep the scope narrow enough to explain. Identify the marketplace, product group, campaign type, currency and reporting source. For a public case study, use permissioned, suitably anonymized information rather than customer account identifiers.

Completed example: compare two equal periods

Every value and campaign detail in this example is fictional. These are not Arctavia customer results. The example uses one UK Sponsored Products campaign, GBP and the same purchase-attribution definition for both periods. Both periods cover 28 days. The question is whether lower spend improves efficiency while retaining attributed sales.

Field	Baseline period	Review period
Start and end dates	1–28 June 2026	29 June–26 July 2026
Example report export time	17 August 2026, 09:00 UTC	17 August 2026, 09:00 UTC
Advertising spend	£500	£400
Attributed sales	£2,000	£1,800
Attributed purchases	40	36
ACoS: spend ÷ attributed sales	25.0%	22.2%
ROAS: attributed sales ÷ spend	4.00	4.50
Total business sales	Not collected in this example	Not collected in this example
Contribution before advertising	Not collected in this example	Not collected in this example
Inventory, price and promotions	In stock; price £50; no promotion	Same conditions
Campaign configuration	Target keyword bid £1.00	Target keyword bid reduced to £0.80 on 29 June

These are illustrative report totals, not a prediction of what a 20% bid reduction will produce. Real reports need their source, export timestamp and metric definitions retained alongside the table. “Not collected” is missing evidence; it does not mean zero.

What the numbers actually show

• Spend falls by £100 (20%).
• Attributed sales fall by £200 (10%) and purchases fall by 4 (10%).
• ROAS rises from 4.00 to 4.50 (12.5%).
• ACoS falls from 25.0% to 22.2%, a decrease of approximately 2.8 percentage points.

A completed conclusion would read: “After the example bid change on 29 June, spend fell from £500 to £400 over an equal 28-day period. ROAS improved from 4.00 to 4.50, while attributed sales fell 10% from £2,000 to £1,800, so sales were not fully retained. We cannot establish a profit improvement or a causal effect from this comparison; next we would review total sales, costs and a comparable campaign without the change.”

Copy the worksheet into your spreadsheet

The table above is a filled example, not an input form. To create your own report, copy the tab-separated worksheet below and paste it into cell A1 in Google Sheets or Excel. Replace the empty cells with your own report values. Keep uncollected metrics marked “Not collected” rather than entering zero. This page does not receive or save your data.

Field	Baseline period	Review period	Source and definition
Period start			
Period end			
Report export time and timezone			
Marketplace / campaign / currency			
Advertising spend			
Attributed sales			
Attributed purchases			
ACoS (spend / attributed sales)			
ROAS (attributed sales / spend)			
Total business sales	Not collected	Not collected	
Contribution before advertising	Not collected	Not collected	
Inventory / price / promotions			
Changes and intervention dates			

Calculate ACoS and ROAS only when the denominator is positive and the source values are available. A percentage calculated from missing values is not a result.

Check competing explanations

Before publishing a conclusion, record whether pricing, stock, promotion timing, seasonality or the campaign mix changed. Keep the same attribution definitions in both periods. Recent conversions can also change an earlier export; see the attribution-window checklist.

A before-and-after report is observational. When feasible, retain a comparable group without the intervention or use a controlled experiment. If you cannot isolate the effect, describe the relationship you observed and the limits of the comparison.

Write the conclusion in three sentences

1. State the change and the reporting period.
2. State the measured primary result and the trade-off, including absolute values.
3. State what the evidence does not establish and what you will review next.

If the objective was acquisition, add new-to-brand sales and purchase metrics with their definitions. Do not turn purchase counts into unique new customers without supporting evidence.

Questions to ask when reading someone else's case study

Is the percentage based on a meaningful baseline?

Ask for starting and ending values, the period length and the sample size. A percentage without those details is difficult to assess.

Were several changes made at once?

If bids, price, inventory and creative changed together, the report may describe the combined result. It cannot automatically assign the gain to one component.

Was profit actually measured?

ROAS is a revenue-to-advertising-cost ratio. Review the cost assumptions separately. Use the break-even ACoS calculator to work through your own economics before adopting another advertiser's target.

Related Arctavia resources

Use these supporting pages to compare Amazon PPC operating models and implementation choices.

• ACoS, ROAS and TACoS: Formulas and Profit-Based Targets
• Amazon Ad Rejections: Check Claims, Images and Destinations
• Amazon Ads ACoS Suddenly Increased: What to Check First
• Amazon Ads ML Bid Review: Guardrails Before Aggression
• Amazon PPC Automation: Data, Decisions and Operating Controls
• Amazon PPC Automation Migration Checklist