Event taxonomy design checklist
Review naming, parameters, duplication, privacy, and QA before release.
Marketing report samples and CSV templates
Each analysis tool expects slightly different columns. Download a blank header-only CSV, fill it in, and upload it without rebuilding the mapping. Downloads are generated in your browser and never sent to a server.
Start with the work you need to deliver
Follow a worked example, check its input requirements, then analyze your own data. Analysis and sample files are free; report downloads require a purchased pass.
Prepare a weekly review
Connect a period comparison to a decision and its next review.
Investigate a CPA increase
Separate spend, conversion and composition changes without asserting causation.
Assess the case for scaling
Distinguish average from marginal performance and hold when evidence is insufficient.
Open a real sample before buying
Free Word and Excel files calculated from the built-in sample. Includes source rows, channel totals and CPA formulas. Includes a chart image in Word and an editable native chart in Excel.
Prepare a file you can reuse next week
- Choose campaign-level reporting and a daily time breakdown. Export a CSV covering both comparison periods; exclude totals and partial days.
- Include the date, campaign, amount spent and the conversion event you will compare. Add impressions, clicks and conversion value when needed. Keep one currency, time zone and attribution setting.
- Check which event the Results column represents. Do not combine leads, purchases and installs as one conversion. Verify column mapping before analysis.
Use Sheets for repeat imports
Keep one input tab with stable columns. Choose a method that matches its sharing permissions.
- Use a sheet already approved for public viewing. Signing into this product does not grant access to a private Google Sheet.
- Open the input tab and copy its full address, including gid when present. Use the normal spreadsheet URL, not a Publish to web link.
- Choose Import from a public Google Sheet in the data input area, paste the link and select Import. Check the column mapping, dates and currency, then run the analysis.
- When the source changes, select Fetch latest data and check the imported period before analyzing again. The browser reads Google directly when you request it; this is not background synchronization.
- Keep the sheet private. In Google Sheets, open the tab containing the rows you want to analyze.
- Choose File → Download → Comma-separated values (.csv, current sheet). CSV downloads contain the selected tab, not every tab in the workbook.
- Upload the file here. Confirm the mapped columns and full comparison period before analyzing. On the next visit, download the updated tab and use the same project.
BigQuery summary table → Connected Sheets extract → CSV upload or public-sheet refresh
- Arrange ingestion into your own BigQuery project first. Scheduled queries transform available data; they do not connect every ad account for you. Google Cloud billing and the required dataset/query permissions must be configured.
- Schedule a query to produce the input table at the grain required by the tool—for example, one day × campaign for campaign performance. Keep currency, time zone and conversion definitions consistent. Schedule it after source ingestion completes.
- On a computer, open Sheets → Data → Data connectors → Connect to BigQuery. Select the project and prepared table or view, then connect.
- Create an Extract with the needed columns and date range, then apply it. Check the extracted row count and period: a preview or a limited extract is not the complete dataset.
- In Refresh options → Scheduled refresh, set a refresh after the BigQuery table is updated. These are separate schedules. Check the last successful refresh; access changes or a paused schedule can leave old data. Query usage may incur Google Cloud charges.
- For company data, keep the extract private and download it as CSV. Only an approved public tab can use this product’s link importer. Either way, import the latest data and run analysis here yourself.
Next time, keep the same columns and measurement definitions and update the dates. Saved settings do not refresh the source: upload the new file and check the comparison periods.
Weekly check
Weekly check
Compare the recent window with the prior one and pick one change.
Budget split
Budget split
Shift until the next unit of spend is equally efficient everywhere.
Change drivers
Change drivers
Split the change into volume, efficiency, and mix.
Headroom check
Headroom check
If marginal CPA far exceeds average CPA, it is already saturated.
A/B testing
A/B testing
Check whether the interval excludes zero, and whether the sample was sufficient.
Paid vs Organic
Paid vs Organic
See whether Paid-up and Organic-down weeks repeat; one or two weeks prove nothing.
Trend
Trend
Separate trend and seasonality first, then read only what is left above that baseline.
Cannibalization
Cannibalization
Four signals check it; even when they line up, causality needs a holdout test.
Channel contribution (MMM)
Channel contribution (MMM)
Adstock and saturation split the panel into channel, base demand, and event shares.
Forecast
Forecast
Forecast-only regression estimates the next window; sealed OOS validation checks it.
Aha moment
Aha moment
Find the action count and window most associated with users who stay.
Incremental lift
Incremental lift
Estimate only the advertising-created share with a holdout or pre/post design.
Brand lift
Brand lift
You cannot pause it, so extend the pre-period trend and compare.
Channel overlap
Channel overlap
If channel spends always moved together, contribution cannot be separated.
ASA keywords
ASA keywords
Promote proven search terms to Exact and adjust their CPT.
Store conversion
Store conversion
If per-source conversion held, the cause is the traffic mix, not the page.
Action survival
Action survival
Estimate survival, hazard, and observed recurring value from censored action episodes.
Composition change
Composition change
Compare two periods' segment shares and split the shift into movement vs internal change.
Multi-touch map
Multi-touch map
Compare recorded click contributors, channel overlap and observed install paths.
Content elements
Content elements
Compare per-element association with a regression that controls the others.
Creative fatigue
Creative fatigue
Retire it when performance drops from its early level while impressions hold.
Make the pre-launch checks repeatable
Download the key QA steps from the operating guides as Markdown, then paste them into Notion, a document, or an issue.
Postback integration QA checklist
Check missing or duplicate events, revenue values, retries, and privacy.
New ad channel onboarding checklist
Fix the objective, measurement, budget, creative, and stop rules before launch.
Frequently asked questions
No. It contains only a header row. Fill the rows below with your own data before uploading.
Yes. Automatic column mapping recognizes most existing headers. Use a template when the mapping is unclear.
No. CSV files are processed in your browser and never sent to a server. Device storage can be managed in Storage.
Download a template below, then use File → Import → Upload in Google Sheets. Keep the header row unchanged, fill your data, then download it as CSV and upload it. Confidential sheets can stay private.
Templates are listed for each published tool with a supported CSV schema. Check the column guide for its required data and observation unit.