Reference (partial) — Workflow and examples below are verified against Admin screens and the tag-rule processor. Adjust event parameters to your brand’s questionnaire ids and status lists.

Recommended workflow

  1. Name the rule — Choose a clear Name; the system creates a Key used when storing tags.
  2. Add conditions — In Tag Conditions, add one row per activity check: pick an Event, set Count and Months, and fill parameters when shown. See the analytics events catalog.
  3. Combine conditions — On Rule Information, choose AND (must meet all rows) or OR (any row is enough).
  4. Optional filters — On Rule Information, you can restrict by profile status, turn on Invert, and set tag Lifetime. See Profile status, invert & lifetime before using invert.
  5. Save and activate — Save a new rule; set Rule status to active when ready.
  6. Link to profiles (optional) — Use Create Attribute so the tag appears as a normal attribute under External Info.
  7. Run — Click Execute to test, or wait for the nightly automatic run. Review Matching Profiles.

Full side effects: Tag Rules

Condition row fields

Every column on the Tag Conditions editable table.

ColumnWhat to enterEffect
Event Pick from the analytics event catalog. Defines what activity is counted. Changing the event clears dependent parameter pickers.
Param 1 Optional. Label changes per event (for example questionnaire name, screen id, login result). Filters the event stream before counting. Leave blank to count all occurrences of that event type.
Param 2 Optional second filter when the event supports it (for example introduction response reaction). Further narrows which events count toward the threshold.
Count Minimum number of matching events required (whole number). Member must meet or exceed this count inside the lookback window.
Months Lookback window in months. Only events within this many months are included in the count.

Parameter labels and dropdown options come from the selected event — see analytics events catalog for param hints per event. Multi-select params store several values; the member qualifies if any selected value matches.

Create Attribute / View Attribute: After save, use header actions to materialize the rule key as an External Info attribute (default group external) so filters and communications can reference the tag like any other profile field.

Matching Profiles: After Execute or the nightly job, open this tab to preview which profiles received the tag under current conditions.

Profile status, invert & lifetime

Optional fields on Rule Information that narrow who gets tagged after event conditions pass.

Field (Admin)What to enterEffect when the rule runs
Profile Statuses (multi-select) Pick one or more lifecycle statuses (for example Approved, Active). Leave empty to ignore status. When at least one status is selected, only members whose current profile status is in that list are considered for tagging. Event counts are still evaluated first; this filter applies to the resulting list.
Invert (toggle) Off by default. Turn on only when Profile Statuses is set. Normal (off): Tag people who matched your event conditions and have one of the selected statuses.
Inverted (on): Tag people who did not match the event conditions, and have one of the selected statuses, and whose account was created more than 30 days ago. Useful for “inactive” or “missing activity” campaigns—test with Execute before relying on the tag in communications.
Tagged Expiration Days (Lifetime) Optional number of days. Leave blank for tags that never auto-expire from this rule. On every manual Execute and nightly automatic run, the system removes stored tags for this rule’s key when the tag row has not been refreshed within that many days. Members who still qualify on the next run get the tag written again.

How invert interacts with event conditions

Event rows (event type, count, months, parameters) are always evaluated against analytics first. Invert does not flip AND/OR on those rows—it changes which profiles receive a tag after the event query completes:

  • Invert off — “Tag everyone who did enough of this activity (and passes profile status, if set).”
  • Invert on — “Tag everyone in the selected statuses who did not appear in the activity results (and joined more than 30 days ago).” Event counts are not written as the tag value in invert mode—the tag marks presence with a simple flag-style value.

If Profile Statuses is empty, invert has no effect in the processor—always pick statuses when using invert.

Lifetime in plain terms

Think of lifetime as “refresh or expire.” A member who keeps qualifying each run keeps the tag. A member who stops qualifying will still show the old tag until the lifetime window passes without a refresh, then the tag is removed automatically. Pair lifetime with communications or reports that read External Info so stale segments do not linger indefinitely.

Example rules

Starting points—confirm ids and statuses in your brand before activating.

Finished onboarding questionnaire

  • Event: Batch finished
  • Param1: your onboarding batch id
  • Count: 1
  • Months: 3
  • Combine: AND (single row)

Tags members who completed that questionnaire at least once in the last three months.

Requested a new introduction

  • Event: Request introduction
  • Count: 2
  • Months: 1

Tags members who tapped “request introduction” (or equivalent) at least twice in the past month—useful for engagement follow-up.

Reached a specific profile status

  • Event: Profile status updated
  • Param1: pick the status from the dropdown (for example “Active”)
  • Count: 1
  • Months: 6

Tags anyone whose status moved to that value at least once in six months. Pair with profile-status filter on the rule if you only want people who are currently in selected statuses.

Two activities required (AND)

  • Row 1: Intro sent — Count 1 — Months 3
  • Row 2: Intro response — Param reaction = declined — Count 1 — Months 3
  • Combine conditions: AND

Tags members who received an introduction and also declined at least once in the same window. See reaction values for other filters (for example semiApproved).

Started a chat after approving

  • Row 1: Intro response — reaction = approved or semiApproved — Count 1 — Months 2
  • Row 2: Chat started — Count 1 — Months 2
  • Combine conditions: AND

Tags members who accepted an introduction and sent the first message in that chat within two months—useful for engagement quality checks.

Skipped a specific questionnaire step

  • Event: Batch content response
  • Param1: content id of the step
  • Param2: skipped = true
  • Count: 1 — Months: 6

Tags members who used skip on that step at least once in six months—not those who answered normally.

Either login or questionnaire activity (OR)

  • Row 1: Login attempt — Count 1 — Months 1
  • Row 2: Batch started — Param1 = your home Lora questionnaire name — Count 1 — Months 1
  • Combine conditions: OR

Tags members who signed in or opened that questionnaire at least once in the past month—useful when you want “any recent engagement” without requiring both actions.

Shared phone in match chat

  • Row 1: Chat started — Count 1 — Months 1
  • Row 2: Media sent — Param mediaType = photo — Count 1 — Months 1
  • Combine conditions: AND

Tags members who opened chat and sent at least one photo in the same month—engagement quality for matched pairs. See chat events.

Inactive actives (Invert)

  • Event: Login attempt
  • Count: 1 — Months: 3
  • Profile statuses: your approved/active status
  • Invert: on

Tags approved members (account older than 30 days) with no login attempts in the last three months. Requires profile statuses—see Profile status, invert & lifetime.

Every selectable event: Analytics events catalog.

Dynamic SQL rules

When condition rows cannot express the question you want to ask.

Condition rows answer exactly one shape of question: "did event E happen at least N times in the last M months?" A rule can instead run a SQL query you describe in plain language, generated by AI and previewed before it ever runs unattended. Use this when you need a ratio, a sequence ("started a chat after approving an intro"), a join against Postgres tables condition rows cannot reach (introductions, profile_responses, support tickets), or anything with more than one comparison operator.

  1. Describe who should be tagged — On the Generate Query tab, pick a source (Postgres or BigQuery events) and describe the criterion in a sentence, e.g. "profiles who declined 5+ intros in the last 6 months."
  2. Review the generated SQL — On the SQL Query tab, read (and edit, if needed) the query the AI produced. It must return rows shaped (profile_id, value) — see Dynamic SQL tag rules for the full contract.
  3. Validate & Preview — Click this before saving. It runs the query capped and read-only, and shows the matched count and a sample of profiles. A rule whose SQL was edited without a fresh preview is skipped by the nightly run rather than running unattended.
  4. Save — Once previewed, save the rule as usual. Matching Profiles, Execute, and tag expiration all work exactly the same as condition-row rules.

A rule is either condition-row-based or dynamic-SQL-based — once a rule has a generated query, the Tag Conditions tab no longer applies to it. Full contract, allowed tables/parameters, and what the validator refuses: Dynamic SQL tag rules.