Measurement
Part of Marketing automation measurement: a reporting framework
Marketing automation key metrics: three reporting cards
Use marketing automation key metrics to define suppression, journey completion and contribution measures, with populations, denominators and decision triggers.
To assess marketing automation reporting, use one measure from each layer: delivery control, audience behaviour and business outcome. These cards assess whether the connected setup can report those layers clearly; add another measure only when it changes a decision.
Marketing automation key metrics should reveal delivery control, audience behaviour and business outcome without mixing their meanings.
The GOV.UK Service Manual recommends defining data needs before collection. That guidance is for government services, but the discipline applies here: write the question, event and action before building a chart.
What to take away
- Use one measure from each layerdelivery control, audience behaviour and business outcome, without mixing their meanings.
- Define the question, event and action before building any chart, following the discipline in the GOV.UK Service Manual.
- Every metric card needs a population, numerator, denominator, unit, period, timezone, source, uncertainty and owner.
- Report contribution per eligible person as descriptive unless an approved experiment supports an incremental claim.
Which platforms already report these layers
Most marketing-automation platforms ship standard delivery and response metrics. HubSpot's email analytics and custom report builder expose open rate, click-through rate, unsubscribe rate and bounce rate. Klaviyo reports campaign and flow performance with deliverability measures and attributed revenue. Mailchimp's campaign reports give open rate, click rate, bounce rate and unsubscribe rate, while its audience reporting covers list growth.
Platforms built for longer journeys report differently. Marketo Engage reports programme performance and pipeline, while Salesforce Marketing Cloud tracks sends and journeys through Email Studio and Journey Builder. ActiveCampaign reports automation and campaign engagement, and Braze adds campaign analytics and Currents for streaming event data. MQL and SQL volumes, conversion rate and campaign-sourced pipeline or ROI usually come from the CRM the platform writes to, not from the sending tool.
Treat those figures as the baseline for a platform review, then test the cards below against them. If a decision depends on suppression propagating across several destinations, the preference centre and connector logs matter more than any built-in campaign report, because no platform dashboard separates eligible population from delivered audience by default.
Metric card 1: suppression application rate
Suppression application rate
- Population
- production records in England that received a valid direct-marketing objection during the reporting window.
- Numerator
- unique people whose objection was effective in every in-scope marketing destination by the approved deadline.
- Denominator
- all unique people in that population, including unresolved and failed propagations.
- Unit, period and timezone
- percentage for a named calendar month using Europe/London event cut-offs.
- Inclusions and exclusions
- include every listed source and destination; exclude test identities and duplicate notices after reconciliation.
- Source
- timestamped preference ledger, connector logs and destination extracts with query version.
- Uncertainty
- identity mismatches, delayed events and missing destination evidence; show unresolved cases separately.
- Owner and update trigger
- data-protection operations owner; revise after a channel, identity rule, deadline or system change.
Illustrative example: if 1,000 production records received a valid objection and 970 were effective by the deadline, the suppression application rate is 97 per cent. The 30 unresolved records stay in the denominator.
The ICO states in its right-to-object guidance that the right is absolute where personal information is used for direct marketing. The law behind that right is UK GDPR Article 21, and PECR regulation 22 governs electronic mail marketing. A UK legal reviewer must set the applicable workflow and timing; the rate does not prove compliance.
Suppression application rate
- Populationproduction records with valid objection
- Numeratorobjections effective in every destination
- Denominatorall unique people, unresolved included
- Unitpercentage per calendar month, Europe/London
Metric card 2: journey completion rate
Journey completion rate
- Population
- eligible records entering one approved production version for the stated England audience.
- Numerator
- unique entrants reaching the defined terminal event once.
- Denominator
- all eligible entrants, including records waiting or failed at period close.
- Unit, period and timezone
- percentage by weekly entry cohort, followed for a fixed number of days in Europe/London time.
- Inclusions and exclusions
- include technical failures; exclude authorised tests and records invalidated before entry.
- Source
- versioned journey log joined to the audience snapshot.
- Uncertainty
- late events, duplicate identities and incomplete connector history; retain a pending category.
- Owner and update trigger
- marketing operations owner; re-version after logic, event-schema, retry or completion-definition changes.
This is an operational outcome. It should not be described as sales conversion.
Journey completion rate
- Populationeligible records entering approved version
- Numeratorunique entrants reaching terminal event once
- Denominatorall eligible entrants, pending included
- Unitpercentage by weekly entry cohort
Metric card 3: contribution per eligible person
- Populationall people eligible for the defined evaluation, whether or not a message was delivered.
- Numeratorreconciled GBP contribution from qualifying orders after stated refunds and variable costs.
- Denominatorunique eligible people under the frozen identity rule.
- Unit, period and timezoneGBP per eligible person for the approved outcome window in Europe/London time.
- Inclusions and exclusionsinclude assigned groups and qualifying orders; exclude unrelated products, periods and unsupported pipeline values.
- Sourcefinance ledger, order records, eligibility file and analysis code version.
- Uncertaintymissing matches, margin assumptions, contamination and sampling variation where relevant.
- Owner and update triggerfinance and evaluation owners; review after a margin, refund, identity or study-design change.
HM Treasury's 2026 Magenta Book distinguishes outcomes from causal attribution. Report this metric as descriptive unless an approved experiment or other credible design supports an incremental claim.
Contribution per eligible person
- Populationall eligible people, delivered or not
- Numeratorreconciled GBP contribution after refunds
- Denominatorunique eligible people, frozen identity rule
- UnitGBP per eligible person, Europe/London window
Set decisions beside the cards
A decision trigger has three parts: the threshold that fires it, the owner who acts and the action taken. State who acts when a control fails, when evidence is inconclusive and when the definition changes. Preserve prior versions and annotate breaks. A rate without its population, denominator and response rule belongs in diagnostic exploration, not an approval pack.
Before you act
- Write the question, event and action before building a chart.
- Name the population, numerator and denominator for each card.
- Record inclusions, exclusions, source and uncertainty for every metric.
- Assign an owner and an update trigger to each card.
- State who acts when a control fails or evidence is inconclusive.
- Preserve prior versions and annotate breaks in definitions.
Common questions
What does the suppression application rate measure?
It measures the share of production records in England with a valid direct-marketing objection that was effective in every in-scope destination by the approved deadline. The denominator covers all unique people in that population, including unresolved and failed propagations.
Why is journey completion rate not sales conversion?
The article describes journey completion rate as an operational outcome, counting unique entrants who reach a defined terminal event once, divided by all eligible entrants. It should not be described as sales conversion. The measure uses weekly entry cohorts followed for a fixed number of days in Europe/London time, with technical failures included.
When can contribution per eligible person support an incremental claim?
Only when an approved experiment or other credible design supports it. Its numerator is reconciled GBP contribution from qualifying orders after stated refunds and variable costs, divided by unique eligible people under the frozen identity rule.
