Measurement Planning
Ecommerce measurement planning
Plan ecommerce measurement around business decisions, clear metric definitions, source records and a review cycle.
Start an ecommerce measurement plan with the decisions the store must make. For each decision, identify the evidence, its source and the person who can act on it. Define the measures before building a dashboard.
List the decisions
Record the decisions made in daily trading and periodic commercial reviews. An order drop may need prompt investigation; a product range decision may need a longer view accounting for reversals and costs.
| Decision | Primary evidence | Context before acting |
|---|---|---|
| Investigate a drop in orders | Qualifying orders in the store record | Checkout activity, stock availability and recent changes |
| Change a product offer | Product sales under a stated definition | Discounts, reversals and product costs |
| Change marketing spend | Store outcomes alongside campaign reports | Spend, attribution settings and customer quality |
These are examples, not universal targets. Keep a measure in the main review when it can inform a likely action.
Give each source a job
Use the commerce order record to establish which orders exist and what happened to them, and state which order statuses qualify. A sales report and a payment record may serve different purposes; define what each source records before relying on it.
Use behaviour analytics to investigate observed shopping actions. Google’s ecommerce guidance describes events for shopping actions; decide which behavioural questions your review needs to answer and verify that the chosen source provides relevant data. A basket addition or checkout start is not an order.
Use ad-platform reports to examine outcomes credited under each platform’s settings. Keep credited conversions and value separate from store sales. Attribution reports can suggest questions to investigate, but they do not by themselves establish how many additional sales advertising caused.
Ecommerce Measurement Sources and Their Roles
- Commerce Order RecordConfirms which orders exist and their status (e.g. completed, refunded). Defines qualifying orders.
- Behaviour Analytics (e.g., GA4)Tracks user actions like basket additions or checkout starts. Not equivalent to an order.
- Ad-Platform Reports (e.g., Google Ads)Shows credited conversions and attributed value. Does not confirm actual sales.
Plan campaign comparisons
Where campaign labels are used, agree what each UTM field represents: source identifies the traffic source, medium the channel, campaign the promotion, term the ad keywords, and content distinguishes links or content within the same ad.
UTM tracking can help compare the same campaign across social platforms and use traffic and conversion results to inform budget allocation. Bitly primarily shortens URLs and provides basic click data, while UTM codes describe traffic source, medium, campaign and content.
Planning a Campaign Comparison with UTM Tracking
- Define UTM fieldsSource: traffic source (e.g., facebook), Medium: channel (e.g., cpc), Campaign: promotion name, Term: keywords, Content: link variant.
- Apply consistent taggingUse the same UTM structure across all platforms for comparable data.
- Compare results in analyticsEvaluate traffic volume and conversion performance by campaign across channels.
- Inform budget allocationUse insights to shift spend toward higher-performing campaigns.
Define measures before comparing them
For every headline number, record its source, unit, date field, reporting time zone, included orders, exclusions and update time. State whether an amount includes discounts, tax, shipping and later reversals. Show the denominator beside every rate.
Platform-defined measures may use different populations or calculations; record the source definition alongside the measure.
Those measures can differ from an internally defined completed-order rate or paid-revenue figure; calculate and label either measure separately. Check for definition breaks when comparing historical rates.
For customer measures, decide how guest and account orders are linked and what counts as a first qualifying order. Keep uncertain identities visible. An unmatched order does not prove that the customer is new.
Key Ecommerce Measures and Their Definitions
- Completed-Order RateNumber of completed orders / total qualifying orders. Must define 'completed' and exclude refunds.
- Paid RevenueRevenue after discounts, tax and shipping. Excludes returns and reversals.
- Customer Acquisition Cost (CAC)Total ad spend / number of new customers acquired. Requires clear definition of 'new'.
Trace the minimum data path
Trace each measure to the records and fields it needs. Purchase analysis may require an order ID, order time, status, item IDs, currency, amounts and later adjustments. Shopping analysis may require consistent item IDs and relevant events.
Google’s ecommerce guidance describes events for purchases and refunds, and recommends setting currency at event level when sending value data. These are implementation examples, not evidence that a store sends correct events.
Assign an owner to missing definitions. If a measure cannot be tied to a source or matched to an order, show that limitation beside it and decide whether the related action should wait.
At planning stage, Google Analytics ecommerce reporting can help quantify popular products and assess the influence of promotions and product placement on revenue. Decide whether those are questions the review needs to answer before choosing its product measures.
Set a review cycle
Set the operating review frequency according to how quickly a signal needs action, and schedule the commercial review when cancellations, refunds and other revisions have had enough time to settle for the decision. Compare like periods and note promotions, stock gaps and site releases. Preserve the basis of provisional figures so later revisions can be explained.
When tools disagree, check whether they count the same action, population, date and amount. Inspect individual records where IDs permit a match, and leave any remaining difference visible.
In this guide
- Choosing store metrics from business decisionsChoose ecommerce metrics by linking each business decision to a counting unit, comparison, owner and action.
- Defining purchase, refund and repeat-customer eventsDefine qualifying purchases, processed refunds and repeat customers with clear order, timing and identity rules.
