A Google Ads report can be numerically correct and still mislead a client. The export may cover the wrong campaign set. A conversion action may have changed halfway through the month. A positive trend may be a partial-period effect. If those conditions stay hidden, polished commentary only makes the mistake easier to repeat.
This Google Ads report template uses six connected sheets. It keeps the reporting contract, conversion definitions, campaign metrics, budget pacing, QA exceptions, and client commentary separate enough to review. The included example is synthetic and uses a fictional advertiser, Northstar Home Services Demo. It contains no customer or account data.
Copy the six-sheet Google Ads report template
The template files are stored as CSV so you can open them in Excel, import them into Google Sheets, or load them into a maintained dataset. The download package should contain these files:
| Sheet | What one row represents | Main job |
|---|---|---|
| Report setup | One client report | Fix the account, period, comparison, time zone, currency, and status |
| Conversion dictionary | One approved conversion action | Record what counts, where it comes from, and which settings apply |
| Campaign performance | One campaign in one report | Preserve source metrics and the comparison period |
| Budget pacing | One budget scope in one report | Compare actual spend with the approved pace |
| QA exceptions | One completed check | Decide whether the report is ready, limited, or blocked |
| Commentary and actions | One evidence-backed statement | Connect an observation to interpretation, action, owner, and review metric |
Use the six-sheet CSV package to follow the reporting workflow below.
You can try the workflow immediately by opening the six CSV files as separate tabs in Excel or Google Sheets. This is the quickest route for testing the fields, formulas, QA gate, and commentary structure before using account data. Keep the six files together so the shared IDs and field definitions remain consistent across tabs.
For repeatable reporting, the files do not have to remain manual exports. GoalfyData can pull from an authorized API, database, or scripted source through a configured refresh workflow, while an MCP-enabled agent can query and update the governed dataset across later sessions. The source connection, credentials, field mapping, and refresh schedule must be configured and verified first; MCP alone does not create an automatic Google Ads connection.
This is a reporting structure, not a performance benchmark. Replace the synthetic rows with approved account data and preserve the field definitions.
The same files can also become a maintained data product. In the worked example below, each CSV was imported into a separate GoalfyData table. The report settings, conversion dictionary, campaign performance, budget pacing, QA exceptions, and commentary remain distinct, while their keys and business rules stay available to the reporting app.
The GoalfyData dataset view keeps table purpose, update mode, primary keys, relationships, governance rules, permission policies, and sharing controls with the asset.
Start with a report contract, not a chart
The report_setup sheet defines the scope that every other sheet must follow. Record the Google Ads account, current and comparison periods, account time zone, currency, conversion-scope version, update time, and readiness state.
These fields solve ordinary but expensive mistakes. A report may compare 20 days with a full month. A manager account export may include a client that should be excluded. A spreadsheet may display dollars while one source account uses another currency. None of those problems can be fixed by changing a chart color.
Use a stable report_id to connect all six sheets. Do not use the client name as the join key because names can change. The account ID should be restricted or redacted in files shared outside the authorized team.
The template uses three readiness states:
ready: every blocking check passed and commentary can be sent.limited: the report is usable, but a documented warning narrows the conclusions.blocked: a scope or data problem could change the main interpretation.
That state belongs in the data, not only in a private message between analysts. It gives the delivery workflow a clear gate.
Freeze the conversion definition before comparing performance
The word "conversions" is not a complete metric definition. A client report also needs to state which actions are included, which source measures them, whether each action is primary or secondary, how it is counted, what value rule applies, and which conversion windows are active.
Google defines a conversion window as the time after an ad interaction during which a conversion can be recorded. Google Ads lets advertisers set different windows for individual conversion actions. A change to that setting affects what can appear in the report, so it must be treated as a definition change rather than a footnote.
The conversion_dictionary sheet records:
conversion_scope_version
conversion_action
category
source
primary_or_secondary
included_in_conversions
count_setting
click_window_days
engaged_view_window_days
view_through_window_days
value_rule
approved_atGoogle Ads asks advertisers to choose the actions they consider valuable, such as a purchase, signup, or call. Its conversion measurement guidance also distinguishes the signals used for campaign measurement and bidding. Your client report should therefore name the approved action set instead of presenting an unexplained total.
Assign a new conversion_scope_version whenever an included action, source, counting method, value rule, or window changes. If the current and comparison periods use different versions, do not publish a simple conversion trend until you have segmented or reconciled the periods.
This article stays within Google Ads. If the client also expects a single business total across Google Ads, Meta, GA4, and a CRM, define separate multi-channel attribution rules before combining results.
Build the performance and budget sheets from controlled exports
Google Ads provides two practical source routes. Its Report Editor can create custom tables and charts with selected dimensions and metrics. Statistics tables can also be filtered, segmented, downloaded, saved, and scheduled in formats that include CSV, XLSX, PDF, and Google Sheets, according to the official reporting instructions.
The live view turns the maintained campaign and budget tables into an interactive report. Users can refresh the page, inspect chart values, and move from the summary into campaign details and QA findings.
Whichever route you use, save the export criteria with the report:
- account and campaign scope
- date and comparison periods
- campaign status filters
- dimensions and segments
- metric columns
- account time zone and currency
- export time
The campaign_performance sheet keeps source metrics at campaign grain. The synthetic example includes cost, impressions, clicks, conversions, conversion value, and previous-period values. Derived metrics should be calculated from totals, not averaged from campaign rates:
CTR = total clicks / total impressions
conversion rate = total conversions / total clicks
CPA = total cost / total conversions
ROAS = total conversion value / total costIf conversion value is based on a fixed lead value, say so. A platform ROAS calculated from assigned lead values is not the same as reconciled revenue from completed sales.
The budget_pacing sheet adds an operational view that most metric summaries omit:
expected spend to date = monthly budget × days elapsed / days in month
pacing variance = spend to date - expected spend to date
projected month-end spend = spend to date / days elapsed × days in monthLinear pacing is a planning reference, not a universal optimization rule. Promotions, weekdays, seasonality, inventory, and campaign learning can justify a different curve. Store the approved pacing rule when the account does not use a linear plan.
The example Dashboard turns those rows into a compact account view. It shows USD 17,600 in spend, 281 reported conversions, a rounded USD 63 CPA, and 4.51x ROAS. The chart compares campaign cost with assigned conversion value, while the pacing panel projects USD 27,280 against a USD 28,000 monthly budget. These are synthetic figures from the downloadable files, not a recommended target.
The first screen keeps the reporting period and synthetic-data label beside the headline metrics, then places campaign movement next to budget pacing.
Open the live Google Ads client report Dashboard. Access depends on the app's current GoalfyData sharing setting.
Run QA before anyone writes the client summary
The qa_exceptions sheet turns review work into records. Each check has a severity, scope, observed state, expected state, owner, status, and resolution note.
At minimum, check:
- The account, date range, comparison period, time zone, and currency match
report_setup. - Campaign filters include the intended enabled, paused, removed, experiment, and newly launched campaigns.
- The conversion-scope version is comparable across periods.
- Cost, clicks, conversions, and conversion value reconcile to the approved Google Ads view within the documented tolerance.
- Partial periods, missing rows, delayed imports, tracking changes, and unusual zeroes are labelled.
- Every client-facing claim points to a stored observation.
The synthetic template contains a blocker: the conversion scope changed on 2026-07-15. It also contains a warning for a campaign launched on 2026-07-10. The account time zone and currency check passed.
A blocker does not make the whole export useless. Spend and click delivery may still be reportable. It does mean the team should stop making direct conversion-efficiency claims until the affected periods are segmented. This is why a report needs status at both report and statement level.
In the example app, the conversion-scope blocker sits beside the affected commentary. The report can still show delivery metrics, but it labels the efficiency interpretation as blocked by QA.
For a broader operating model around ownership, delivery, and approvals, use the client reporting system alongside this channel template.
Write commentary as evidence, interpretation, and action
Client commentary becomes easier to review when it is stored in parts. The commentary_actions sheet uses this sequence:
evidence observation -> interpretation -> confidence -> proposed action
-> owner -> due date -> client decision -> review metricThe evidence should be a statement the report can reproduce. The interpretation explains what the observation may mean. The action states what will change or what will be investigated. Confidence prevents a tentative reading from becoming a certain claim.
Weak commentary:
Performance Max improved, so we should increase budget.
Reviewable commentary:
Reported Performance Max conversions increased from 75 to 82 in the synthetic example. The approved conversion scope changed on 2026-07-15, so the increase cannot yet be attributed to campaign improvement. Segment results around the change and do not raise budget from this comparison alone.
The second version is longer because the decision has a real constraint. It names the observation, limits the interpretation, and identifies the next check.
Another synthetic row shows Nonbrand Search cost rising from USD 7,400 to USD 7,800 while reported conversions fall from 115 to 101. That would normally prompt a search-term and landing-page review. Because the conversion scope changed, the template marks the interpretation as low confidence and blocks the client claim until reconciliation is complete.
Do not hide that condition in a footnote. The client needs to know whether the team is recommending a budget change, requesting a decision, or investigating a measurement issue.
Turn the six sheets into a monthly client report
Once blocking checks pass, the client-facing report can stay compact:
1. Scope and status
Show the period, comparison, account scope, currency, conversion definition, and report status. Include any limitation that changes how the results should be read.
2. Outcome and delivery summary
Present spend, conversions, CPA, conversion value, and ROAS only when their definitions are approved. Keep impressions, clicks, CTR, and CPC available as delivery context rather than treating them as business outcomes.
3. Campaign movements
Choose the few campaign changes that explain the account movement or require a decision. Avoid a long winners-and-losers table with no action attached.
4. Budget pacing
Show spend to date, approved budget, pacing variance, projected spend, and the next review date. Explain any non-linear budget rule.
5. Exceptions and next actions
List open limitations, agreed actions, owners, due dates, client decisions, and review metrics. A resolved exception can remain in the record without taking over the client summary.
This structure also fits a recurring weekly reporting workflow. The report frequency can change while the contract, QA checks, and action history remain stable.
Maintain the template without losing its rules
In Excel or Google Sheets, place each CSV on a separate tab and protect stable identifiers and formula columns. Add a README tab that records the source export steps, tolerances, owners, and change history. Do not overwrite the previous report when a new period starts.
A database or governed dataset can keep the same six-table structure. The useful part is not the storage choice. It is the ability to preserve definitions, relationships, approval state, and past actions while new exports arrive.
GoalfyData can store approved agent output as structured datasets with field context, rules, relationships, and controlled access. In this example, the six tables are a reusable asset rather than data embedded in a single Dashboard build. The reporting app reads the maintained dataset, so a reviewed data update can flow into the client view without rebuilding the reporting logic around another spreadsheet copy.
An agent can help load a new Google Ads export, run the documented checks, and update the reporting app, while the team keeps the contract, QA rules, and table relationships available across later sessions. This article does not assume a native Google Ads connector or an automatically refreshed account. The example uses a manual, one-time CSV import; scheduled refresh would require a separately configured and verified data source.
Share the data and the client view at the right scope
A report should not become easier to share by becoming harder to govern. GoalfyData separates access to the underlying dataset from access to the deployed reporting app.
For dataset collaboration, an owner can send private email invitations or create a reusable public code with an explicit read-only scope. The owner can share all data, limit access to selected tables, or apply a permission policy that filters rows and columns. That makes it possible to keep internal QA or account identifiers out of a recipient's view instead of maintaining a second uncontrolled workbook.
The sharing dialog offers all-data access, selected-table access, or a permission policy for row- and column-level restrictions.
The reporting app has its own sharing path. It can remain owner-only, be shared with specified recipients, or be published through a managed public link when the report is intended for broad distribution. The worked example linked above currently requires GoalfyAI sign-in. Before sharing a real client report, confirm the recipient list, expiry, visible fields, and whether the client needs the interactive app or only an approved export.
The final gate remains human: confirm the conversion scope, review exceptions, approve the commentary, and record the decision. A reusable template reduces repeated assembly. It should not remove accountability from the report.
