Skip to content

Measure what happened after activation

An outcome defines what success means after an audience is activated. CAB follows the activated records, takes readings at scheduled checkpoints, and keeps those results separate from the live Salesforce records behind them.

Open CAB Outcomes to find your outcomes, then select a name to open its detail page. Screenshots use synthetic donation data and show the English interface.

The library puts measured outcomes first. Search by name or sort by Rate or Name. A measured zero belongs with the measured results; it is different from an outcome that has not started.

The Outcomes library showing five measured results, including a zero, a volunteer outcome waiting for its first reading, and three outcomes that are not measuring yet.

Use the status to decide what to inspect next:

  • A rate is available: read the checkpoint and measurement date. A measured zero means the reading completed without qualifying conversions.
  • The first reading is pending: activation has started, but CAB has not taken a reading yet. Check the activity timeline for the next measurement.
  • Nothing has been counted: review the action assignment and its schedule. An unassigned outcome needs Assign to an action; an assigned outcome needs an action that will run.
  • All readings are complete: the result remains available even when no future activation is scheduled. A later activation starts another population.

The detail page brings the result, measurement activity, and analysis together. Start with the summary at the top, then use the timeline and tabs to investigate it.

The complete outcome detail page: summary and actions, measurement activity timeline, three record tabs, audience ranking, monthly metrics, and How this is measured.

The header distinguishes Latest reading from All months together. Read the month and checkpoint beside the rate: “day 7” is an early reading, even when the conversion window lasts 90 days.

The activity timeline shows completed measurements, upcoming readings for records already tracked, and the next audience run. A first-reading date in a monthly chart can be later than an individual activation’s next checkpoint: a whole activation month must be old enough for that reading.

Use the three record tabs:

  • Measurements for results, comparisons, and charts.
  • Tracked activations for the populations captured by each activation.
  • Definition for the saved success rule, conversion date, window, checkpoint days, and assigned actions.

Choose the metric that answers your question

Section titled “Choose the metric that answers your question”

Under Metrics, the tabs answer different questions. They do not merely change the chart style.

Metric Horizontal axis and meaning
Conversion rate Activation month. Of the records activated in that month, what proportion converted by the stated checkpoint? Gray bars show the number activated. The audience tracks below share the same calendar.
Conversions Conversion month. How many conversions happened in each month, according to the configured conversion date? Read the caption to see whether the unit is unique people or activation episodes.
Time to convert Days after activation. Which intervals contain the observed conversions? Only activation months that have reached the final checkpoint contribute, so each interval uses the same fully followed population.

Conversion rate across six activation months, with activation-volume bars and audience activity tracks on the same March-to-September axis.

Compare monthly rates at the same checkpoint. Here, every month from March through August has activations and a day-7 reading. The bars show how many records were activated; the tracks show which audiences were active.

Months without readings stay empty. They do not represent a zero conversion rate. A new all-time chart reserves at least six months; longer histories extend the axis, and bounded reading windows retain empty months within the selected range.

Conversions by conversion month from March through September, with several completed monthly counts and a hollow September mark for a month still filling.

Use conversion counts to see when results arrived. Activations in March can produce conversions in April, so the shape differs from the activation-month chart. The caption identifies the counting unit; the hollow September mark shows that its count can still rise.

Time to convert shows a horizontal stacked bar. Each segment represents the share of observed conversions in a checkpoint interval. The legend names the intervals and their shares; the caption gives the fully followed population and its final conversion rate.

Time to convert for a fully followed population of 691 accounts: a stacked distribution across checkpoint intervals, with 125 conversions by the final day-90 check.

Read segment percentages as shares of the 125 conversions, not shares of all 691 activated accounts. The final rate is 125 / 691, or 18.1%. Keeping the population fixed across intervals prevents younger activations from distorting the later intervals.

The distribution is derived from checkpoint observations, not a member-by-member event timeline. Late-entered data can affect when a conversion is first observed.

A hollow conversion mark means that month can still gain conversions. Missing retained timing history is disclosed rather than reconstructed as zero. For Time to convert, CAB explains when no activation month has reached the final check, no one converted, or the outcome has no dated conversion to analyze. Waiting for that distribution does not invalidate an earlier day-7 rate.

The audience ranking holds the checkpoint constant. Select an audience to show it on the chart, select it again to remove it, or return to Everyone · average. Use the full ranking to see uncertainty: it orders audiences by the lower bound of the 95% interval. Audiences below the displayed minimum sample size are counted but not ranked.

Impact of activation compares Activated records with Held back for comparison when a compatible control group exists. An organization average is not a control group. See Control groups and lift for how to interpret the difference and its uncertainty.

Expand How this is measured for the definitions that apply to the selected metric. Read notices about incomplete coverage, late observations, small samples, and rule changes before interpreting a trend.

  1. In CAB Outcomes, select New outcome. Choose Campaign membership to read the campaign membership created by the action, or Conversion event to recognize a qualifying Salesforce record. The kind is fixed after creation.
  2. Enter Name and choose Tracked object, matching the record type your audience activates.
  3. Define success. For campaign membership, set the responding flag, status, or other membership criteria. For a conversion event, choose a related record, the tracked record itself, or a parent record it belongs to, then define the filters and relationship.
  4. For a conversion event, choose Conversion date field and Conversion window (days). For campaign membership, set Observation horizon (days). CAB derives and previews the reading schedule from that duration.
  5. Under Listening scope, choose compatible Scheduled campaign actions. You can save an unassigned definition, but it will not measure anything until it is assigned.
  6. Select Save outcome.

For example, to measure whether prospects start giving monthly, activate accounts that might become recurring donors. Define success as an active Gift Commitment related to the account, dated by its effective start date within 90 days. The success rule is separate from the audience’s selection rule: requiring an active commitment in the original audience would select people who had already succeeded.

To measure event attendance, define success as a qualifying campaign-membership status. This kind can measure the membership an action created; it cannot measure campaign response for people who were held back and never added.

When a parent record defines success, every tracked member linked to that record can qualify. For example, two contacts linked to the same account can both count when that account meets the rule. Choose the tracked object to match the unit you intend to measure.

Supported action editors expose Tracked outcomes. Assign the outcome to the action that actually activates the audience. Campaign membership outcomes work with campaign additions; compatible Salesforce conversion outcomes can also be attached to supported Marketing Cloud, Pardot, and custom final actions.

If a reusable conversion rule has parameters, supply the required values where the outcome is used. Those values are part of the configuration being measured. Review the saved Definition before comparing results or using them for a follow-up audience.

Creating or assigning an outcome does not backfill earlier activations. Editing its rule creates a new version; records already captured keep the version they started with.

  • Activation starts the clock. Saving a definition or scheduling an action does not.
  • The conversion window decides which event dates count. For a 90-day window, an event outside that window is not made eligible by a later reading.
  • Checkpoints decide when CAB looks. A 90-day conversion outcome normally checks at days 7, 14, 30, 60, 90, and 120. The final check catches eligible conversions entered late; it does not extend the 90-day window.
  • Campaign membership uses its observation horizon. It has no extra conversion late-entry check beyond that horizon.
  • A stored checkpoint is a snapshot. An eligible event entered after an earlier reading can appear in a later one without rewriting the earlier reading.

The normal editor derives the schedule; it can include more than six readings for a longer window. An existing custom schedule is identified as such. Follow the schedule shown for the definition. Windows and checkpoint days cannot exceed 730 days.

Open Tracked activations on the outcome record. Filter by Audience, Destination, or Activated, then open the activation’s member view. Its checkpoint, member count, converted count, and rate identify the population you are inspecting.

The audience’s Outcomes view and an expanded addition in Run History also provide activation details. Holdout members have a separate View / export control group view; they are not included in the activated-member roster.

In the activation viewer:

  1. Choose a completed Checkpoint. The viewer starts with the latest one.
  2. Switch between All and Converted when available.
  3. Follow Evidence to an accessible qualifying Salesforce record. Compare Evidence date with the outcome’s configured conversion date and window.

A measured monthly-giving activation filtered to converted members, with checkpoint selection and links to the actual Gift Commitment evidence.

After the final checkpoint, the member population remains accessible only for the configured retention period. With the default of zero days, it becomes unavailable when measurement finishes. Aggregate results and conversion-timing history can remain after those individual details are released.

Choose Act on it on the outcome record:

  • Suppress people who already completed excludes people who have already qualified.
  • Follow up with everyone who didn’t complete starts from measured non-completers and protects control-group members automatically.
  • Find more people like the ones who completed uses the available completed population for a similar-audience workflow, when that option is available.

Act on Monthly giving started, offering the available suppression and follow-up choices for its measured population.

For suppression or follow-up, choose Start a new audience or Add to an existing audience, then Open in builder. Review the resulting audience before saving or activating it.

A historical rate alone cannot recreate a member list. CAB explains when the required measured population is absent or has been released. For parameterized outcomes, provide the required configuration values first. Administrators can extend retention for populations that are still stored; increasing it cannot restore records already deleted.