Skip to content
Optima docs
ConsoleAgentsBuilt-in agents

Weekly campaign digest

A built-in agent that emails a weekly summary of how your campaigns and journeys performed.

The weekly campaign digest is a built-in agent that tells your team how campaigns and journeys did over the last week, without anyone opening each campaign. Once a week it reads the results and sends one plain-text email.

Who uses it

Campaign managers and tenant administrators create the instance. The people who receive the email can be anyone on your team or any email address you enter. You need the Built-in agents feature in your licence and the Campaign package with its campaigns, journeys and attribution features. An account that excludes journeys or attribution does not see this agent.

Fields

ParameterWhat you enterDefault
recipientsThe people who get the email.Owner (you)
scheduleWhen the digest is written, as a cron expression in the instance time zone.0 15 * * 5 (Friday, 15:00)
period_daysHow many days back from the schedule time the digest covers. A whole number from 1 to 365; any other value falls back to a week and the email says so.7
campaignsThe campaigns to include. Leave empty to include every campaign that was active in the period.Any
conversion_eventThe event type that counts as a conversion. Leave empty to skip the attribution part.Empty
always_sendSend an email even when no campaign was active in the period.Off

Actions

  • Reads: which campaigns sent messages in the period, their results per step, the control-group lift, the results of one journey campaign and attribution for your conversion event. It reads aggregates, not individual customers.
  • Sends: one plain-text email to the recipients you choose. The subject names the number of days and the date the digest was written, for example "Campaign digest: 7 days to 2026-10-30".

The email contains, in this order:

  • Overview: the campaigns and journeys it covers, by name.
  • What went well: up to three campaigns with the best rate on one step (opened, otherwise clicked), each with its own value.
  • What did not go well: up to three campaigns with the weakest rate on the same step, and the journey steps where the most runs exited or failed. Journey counts are since the journey started, not for the period.
  • Control-group lift: for each campaign with a control group whose lift it read, the lift as Optima returned it.
  • Split tests: tests decided in the period, with their winner.
  • Attribution: the credit under each model, only when you entered a conversion event.
  • Recommendations: at most three, each tied to a figure in the email.
  • What I could not check: failed reads, skipped campaigns and missing data.

When no campaign was active in the period, no email is sent unless Always send is on; then you get a short note saying there was no campaign activity. If the agent cannot read which campaigns were active, it emails that the digest could not be built and why, instead of staying silent.

Every figure in the email is copied from Optima's own results, never estimated, and is checked against what the agent read. If the agent cannot back a figure up, the run fails and nothing is sent.

Limits

The agent covers at most eight campaigns per digest, the ones that sent the most messages, and the journeys of one journey campaign (at most three journeys). "What I could not check" says that the smaller campaigns were left out and names up to five of them. A campaign the agent could not look up by name appears by its id. If you have more active campaigns, create several instances and choose different campaigns in each.

Examples

To try it, create the instance in dry run with a cheap model and set schedule to a time a few minutes ahead. Open the run log to see the email it would send. When you are happy with it, set the schedule back to Friday 15:00 and switch the instance to live.

Common problems

  • No email this week. No campaign was active in the period, or none is in the campaigns list you chose, and Always send is off. Or the instance is stopped.
  • The email has no attribution part. The conversion_event is empty.
  • Some campaigns are listed as not checked. More than eight campaigns were active, or a read failed. Choose the campaigns you care about, or create more instances.
  • The run failed with "numbers that no tool returned". Try a stronger model or look at the run in the run log.
  • The card is missing. Your licence has no Built-in agents feature, or no campaigns, journeys or attribution feature.

On this page