Measuring Marketing Performance

Marketing performance should connect spend and communication to completed customer journeys, contribution and operational consequences. Dashboards that stop at impressions, clicks or app installs can reward activity that never becomes a sustainable order.

Build a measurement hierarchy

Reconcile the inputs

Match campaign spend to invoices, not only a platform dashboard. Confirm code cost, free items, delivery subsidy, agency fees and internal time. Exclude test, cancelled, duplicate and fraudulent orders consistently.

Use an appropriate baseline

Compare with a similar period, audience or holdout. Note weather, holidays, menu changes, store closures and competitor events. A simple year-on-year comparison can mislead if the service area or prices changed.

Protect the operation

Add capacity and quality limits to the campaign report. If Friday delivery orders rose but late deliveries and refunds rose faster, the campaign may have destroyed value. Channel managers should not optimise independently of kitchen and dispatch.

Review owned marketing responsibly

• Audience provenance and valid preference
• Suppression applied at send time
• Complaints, opt-outs and failed delivery
• Frequency across email, SMS and push
• Incremental contribution rather than gross revenue

Use a decision record

At the end of the campaign, choose: scale, repeat unchanged, repeat with one defined change, pause or retire. Record why. Avoid the common outcome where every campaign is called “successful” and repeated without a threshold.

Core scorecard

• Full campaign cost
• Incremental completed orders or best supported estimate
• Contribution after campaign cost
• First-to-second-order rate
• Refund and complaint rate
• Opt-out and suppression indicators
• Kitchen and delivery capacity impact
• Confidence and known data gaps

Practical next step: replace one marketing dashboard with a single-page decision scorecard. If a metric does not change a decision, remove it or move it to a diagnostic appendix.

Set reporting cadence by decision

Operational alerts may need daily review, while acquisition and repeat-order conclusions often need several weeks. Avoid changing campaigns every day based on small fluctuations. Define which metrics are monitored, which trigger action and which are reviewed only after the measurement window closes.

Assign ownership

Marketing owns campaign setup and spend; operations owns capacity and service; finance validates cost and contribution; data owners validate definitions and permissions. A final decision should record all four views rather than allowing one dashboard to overrule the others.

Archive the final creative, audience rule, conditions and report together. This makes future comparisons possible and prevents the same failed experiment being repeated under another name.

Where results are uncertain, state the uncertainty plainly. A decision range with known gaps is more useful than an exact return figure built from incompatible attribution and finance data.

Related guides

• Restaurant Marketing and Customer Retention — the main guide for the wider topic.
• Customer Lifetime Value for Takeaways — a detailed next step for putting this guidance into practice.
• How to Measure Repeat Order Rate — a detailed next step for putting this guidance into practice.
• Loyalty and Retention Strategy — a closely related operational control to review alongside this page.
• Promotional Codes and Discounts — a closely related operational control to review alongside this page.
• Online Ordering Data and Analytics — a connected decision that can change the recommended approach.

Sources and date checked

Guidance checked: 24 July 2026. Recheck official guidance and supplier rules before launch.

• ICO: plan direct marketing
• ICO: storage and access technologies guidance