Skip to content
← ResourcesMeasurement & AttributionMarch 18, 2026 · 6 min read

Find the weakest layer of your measurement setup.

Assess tracking, attribution, reporting and incrementality; select evidence and owners for the next improvement.

The working sequence.

10 sections

Follow the document’s sequence, or jump to the step you need.

  1. 01Layer 1: Tracking Foundation (30 points max)
  2. 02Layer 2: Attribution Clarity (24 points max)
  3. 03Layer 3: Reporting and Diagnostics (24 points max)
  4. 04Layer 4: Incrementality (12 points max)
6 min read13 reference tablesFree · no email

How the scoring model is weighted.

Maximum available points in each area. Use the full assessment below to calculate your own score.

Tracking Foundation30 points
Attribution Clarity24 points
Reporting and Diagnostics24 points
Incrementality12 points

0–30 · available points per area · 90 points in total

Source: The scoring sections in this resource · These are rubric weights, not your score.

Jump to a section

Put this guide to work

Assess tracking, attribution, reporting and incrementality; select evidence and owners for the next improvement.

Your task: Assess your measurement layers. Record the source, date range and reviewer beside the result. Leave anything you have not checked marked Not checked.

How to interpret it: The score summarizes your rubric answers. It cannot certify tracking accuracy or prove incrementality. Confirm each layer with observed evidence; cost-of-gap figures are scenarios rather than measured losses.

Related work: Check the foundations underneath your tracking. · Design a test that can answer an incrementality question.

How to Use This Assessment

Score each item on a 0-3 scale:

  • 0 = Checked and not implemented
  • Not checked = Unknown; leave outside the scored denominator
  • N/A = Not applicable to this setup; exclude from the scored denominator
  • 1 = Partially implemented / Needs work
  • 2 = Implemented but not verified recently
  • 3 = Fully implemented and verified within the last 30 days

Record the evidence beside each rating. A checked gap differs from missing evidence. This rubric cannot estimate misallocated spend from a score. The printed layer maxima assume every criterion applies and is rated: for partial coverage, use 3 × rated applicable criteria, show the rated/total counts and keep unreviewed coverage visible.


Layer 1: Tracking Foundation (30 points max)

GTM Container Health (9 points)

GTM Container Health (9 points)
# Item Score (0-3)
1 Consistent naming convention on all tags, triggers, and variables ___
2 Click-data storage verified through the Google tag or applicable conversion-linker implementation ___
3 Conversion values/parameters match the chosen business definition; any fixed values are intentional and verified ___

Conversion Action Configuration (9 points)

Conversion Action Configuration (9 points)
# Item Score (0-3)
4 Primary conversion goals match the business objective; duplicate or secondary events do not inflate the selected outcome ___
5 Other actions have primary/secondary roles that match the selected objective without double-counting outcomes ___
6 Transaction ID deduplication active (thank-you page refresh creates 1 conversion, not 4) ___

Server-Side and Data Recovery (12 points)

Server-Side and Data Recovery (12 points)
# Item Score (0-3)
7 Enhanced-conversion setup and reported matching coverage reviewed with their limitations ___
8 The selected browser/server implementation meets documented needs; server-side infrastructure is not automatically required ___
9 Selected basic/advanced consent implementation verified against actual choices and requirements; four relevant consent types reviewed ___
10 Cross-domain tracking verified (if separate checkout domain) ___

Layer 1 Score: ___ / 30


Layer 2: Attribution Clarity (24 points max)

Model Configuration (9 points)

Model Configuration (9 points)
# Item Score (0-3)
11 Attribution model and current provider eligibility checked against the actual conversion setup and volume ___
12 Attribution windows explicitly reviewed against the purchase cycle; defaults may remain if appropriate ___
13 Applicable model comparisons reviewed with current availability and scope; unavailable comparisons remain Not checked or N/A ___

Cross-Platform Reconciliation (9 points)

Cross-Platform Reconciliation (9 points)
# Item Score (0-3)
14 GA4 and commerce revenue reconciled using compatible windows, attribution, refunds, tax and currency definitions ___
15 Google Ads/GA4 differences explained using compatible definitions, windows and known coverage; no universal gap band ___
16 A reconciliation dashboard exists showing all sources side by side ___

Conversion Quality (6 points)

Conversion Quality (6 points)
# Item Score (0-3)
17 Conversion delay measured (you know your 80th percentile purchase lag) ___
18 Outcome evaluation accounts for conversion delay; urgent budget, tracking or policy defects can still require immediate containment ___

Layer 2 Score: ___ / 24


Layer 3: Reporting and Diagnostics (24 points max)

Metric Classification (9 points)

Metric Classification (9 points)
# Item Score (0-3)
19 Metrics classified into vanity, diagnostic, and decision tiers ___
20 Dashboard separates decisions, diagnostics, platform ratios and observed contribution/cost coverage ___
21 Alerts configured for diagnostic metrics that breach thresholds ___

Weekly Diagnostic (9 points)

Weekly Diagnostic (9 points)
# Item Score (0-3)
22 Weekly tracking integrity check (conversion count vs 4-week average) ___
23 Budget pacing reviewed weekly (constrained winners identified) ___
24 A named review cadence fits volume, risk and conversion delay; timeboxes are operational choices ___

Monthly and Recovery (6 points)

Monthly and Recovery (6 points)
# Item Score (0-3)
25 Monthly deep dive with budget reallocation review ___
26 Recovery playbooks documented for common crises (CPC spike, ROAS collapse, volume drop) ___

Layer 3 Score: ___ / 24


Layer 4: Incrementality (12 points max)

Layer 4: Incrementality (12 points max)
# Item Score (0-3)
27 At least one incrementality test run in the last 12 months ___
28 Any brand-search replacement estimate retains its actual test population/window and uncertainty; unsupported estimates stay unknown ___
29 Any incrementality adjustment retains valid test scope, uncertainty and review date; no universal factor is applied to daily ratios ___
30 Quarterly testing calendar planned ___

Layer 4 Score: ___ / 12


Your Total Score

Your Total Score
Layer Your Score Max %
Tracking Foundation ___ 30 ___%
Attribution Clarity ___ 24 ___%
Reporting & Diagnostics ___ 24 ___%
Incrementality ___ 12 ___%
TOTAL ___ 90 **___% **

Maturity Levels

Maturity Levels
Rubric score Review priority Evidence to collect
0–22 / 90 Review foundations Actual events, identifiers and source coverage
23–44 / 90 Review configuration gaps Conversion definitions and reconciliation
45–66 / 90 Review reporting and interpretation Consistent windows, attribution and unknowns
67–80 / 90 Review validation Test design, uncertainty and contamination
81–90 / 90 Maintain and verify Observed checks and a named review cadence

The score cannot prove a layer works. Keep evidence and unknowns with each answer. The 0–90 bands apply only to a fully rated 30-item worksheet. For partial coverage, report earned points / rated maximum with coverage and the outstanding evidence; do not assign a full-assessment band.

Cost-of-gap worksheet

Use your own observed discrepancy, contribution and cost data. Do not infer a dollar loss from a failed rubric item.

Cost-of-gap worksheet
Gap Observed evidence and period Assumptions to verify Owner / next check
Duplicate event or conversion credit ___ ___ ___
Missing identifier or value ___ ___ ___
Unreconciled reporting ___ ___ ___
Inconclusive test ___ ___ ___

What to Do Next

For the complete team-operated workflow, review the Google Ads AI Agentic System. The product page lists the current deliverables, price, license terms and setup responsibilities.

For the next review: Choose the lowest-rated applicable layer and collect its missing evidence.

For maintenance: Record a review cadence and recheck after platform, account or team changes.

Use the review to choose the next evidence check. A configuration gap or unknown answer is not itself an observed financial loss.

Choose who handles the next step.

Use the reference yourself, book a focused paid working session, or have Tegra handle the ongoing channel work.

Search, Shopping, Performance Max and Demand Gen campaigns, plus the research, product feeds, static and video creative, and landing, offer and pre-sale pages behind them.

Explore managed Google Ads → · See paid working sessions →
Get the operator playbook · weekly

One tactical post a week. No fluff.

By subscribing, you agree to get Tegra tactics and product news by email. Unsubscribe anytime. Privacy Policy

Operate this workflow with your team.

Google Ads AI Agentic System · $19,997

The licensed system gives your team the source, workflows and references to prepare, review and operate this work. Installation and process-integration help are included at no additional fee, with rollout scope agreed together. Provider usage and ongoing management are separate.

See the system→