Audience Analysis use cases analyze-audiences-use-cases
Audience Analysis enables the reporting of Experience Platform audience membership data in Customer Journey Analytics. This is achieved through configurations managed via the Audience Analysis configurations wizard, which helps you determine which profile dataset you are ingesting, alongside other parameters and configuration details. (For more detailed overview information, see Audience Analysis overview. )
This document includes example use cases that highlight the value that Audience Analysis provides. Before reviewing the use cases, first become familiar with the reporting considerations below. It is important to keep these considerations in mind when going over your use cases, as they may impact the final output of your reports.
Reporting considerations
The initial release of Audience Analysis establishes essential groundwork required for processing and ingesting Experience Platform audiences into Customer Journey Analytics. When conducting your analysis, it is important to be mindful of several factors that may influence your results across Workspace projects:
-
Audience membership data is accurate only for the previous day (“yesterday”): Audience membership data will always contain the latest profile snapshot dataset generated by Unified Profile Service. This profile dataset is a daily snapshot and is accurate only for the previous day (“yesterday”), with it being automatically regenerated and reprocessed each night. Audience dimensions are available for reporting and breakdowns, not for reconstructing historical audience states.
-
Example: Regardless of the chosen reporting time window, the CJA reportable audience will always respect the audience membership state present in the latest ingested profile snapshot (“yesterday”).
- Widening the reporting time window to “last 30 days”, as an example, will include more events and give the impression that the audience size is changing. However, the audience’s profile composition will always match the snapshot of “yesterday” regardless of the chosen time window.
-
-
Dimensions must have a corresponding event to be included: Audience Analysis dimensions can only be analyzed where corresponding events exist in CJA. If a behavior, channel, or lifecycle moment is not represented as an event in the CJA connection, it cannot be analyzed.
- Example: An audience that is used to target people with an ad would include significantly more people in the RTCDP audience than in the CJA audience. This is because the CJA audience is limited to people who had an event in CJA during the reporting window.
-
Identity resolution is based solely on a single namespace: Identity resolution depends entirely on the selected identity namespace as part of the Audience Analysis configuration. Analysis will be constrained to said identity namespace, with events falling outside of it not being available for audience analysis reporting.
- Example: For a stitched event dataset that combines CRM and ECID, and the Audience Analysis configuration uses the CRM ID, only rows containing a CRM ID will be recognized as part of the reportable audience in CJA. Therefore, the resulting audience size may be smaller than anticipated.
Example use cases
Use case 1
Understand how a specific audience behaves in a given channel (e.g., web or app), to answer questions like:
-
“What are members of High‑Value Prospects doing on the site right now (pages, features, funnels)?”
-
“Which campaigns and content over‑index for this audience compared to everyone else?”
Configuration flow
-
Configure Audience Analysis in CJA for a single identity namespace (e.g., ECID or CRM_ID) and a web‑focused data view.
-
This will automatically ingest the audience membership data into your chosen connection via daily profile snapshot export
-
It is recommended that you select the identity namespace that you believe has the most coverage in your event dataset
-
-
Build your Workspace projects to:
-
Break down Audience Name by page, product, campaign, device, etc.
-
Compare audience vs. non‑audience (or against another audience) on metrics like sessions, conversion rate, revenue per person.
-
-
Use the generated insights to fine-tune channel optimization strategies (e.g., targeting rules, content or offer tuning).
Identity resolution considerations
Most activity is already tied to one login/CRM ID, so audiences from AEP join cleanly to behavioral data in CJA.
You get a clear view of what each audience is doing in a given channel with minimal expected reporting gaps.
Even when joined with a stitched dataset, reporting will still be constrained to the single identity namespace chosen in the configuration.
If some customers exist under other IDs, their behavior won’t be included, which may result in a partial view during reporting.
Use case 2
Help marketers or journey designers understand which audiences overlap, so they can de‑duplicate experiences and avoid offer collisions across campaigns:
-
“How much overlap is there today between Loyalty Members, Cart Abandoners, and High Propensity to Churn?”
-
“Which audiences are most likely to collide on high‑value promotional offers this week?”
Configuration flow
-
Configure Audience Analysis in CJA for a single identity namespace aligned to RTCDP/AJO activation (e.g., CRM_ID if journeys are people‑based).
-
This will automatically ingest the audience membership data into your chosen connection via daily profile snapshot export.
-
This can be used to enrich an existing data view that already powers key engagement and conversion reporting.
-
-
Use the out‑of‑the‑box Audience Analysis Overview template and overlap visualizations:
-
Run audience intersections and over/under‑indexing views (e.g., % of Cart Abandoners who are also in Loyalty Gold).
-
Slice overlaps by core dimensions (e.g. device type, product interest) to understand where conflicts matter most.
-
-
Use the insights to fine tune critical aspects, such as:
-
Offer conflict rules or marketing action rules in RTCDP and AJO.
-
Audience refinement (e.g., tightening target definitions where overlap is too high).
-
Identity resolution considerations
Because most activity is already tied to a single login/CRM ID, this is expected to provide a reliable overlap map across audiences.
Overlap charts give a reliable picture of which audiences collide and where to apply suppressions or priority rules.
If parts of the journey live under other IDs (e.g., anonymous ECID‑only browsing, call‑center IDs), those events won’t show up in overlap analysis.
People may still exist across multiple namespaces.
Overlap will be based on the identity namespace included in the configuration. If some profiles still split across IDs, overlap may under‑report true collisions.
Use case 3
Understand the behavior of customers who recently left a key audience and what they did around that exit, to answer questions like:
-
“Who just left a key audience, and what did they do around exit?”
-
“What happened right before exiting? (errors, low engagement, price changes).”
Configuration flow
-
Configure Audience Analysis in CJA for a single identity namespace (e.g., CRM_ID or login ID) and the relevant data views (web, app, CRM, etc.).
-
This automatically ingests the audience membership data into your chosen connection via daily profile snapshot export.
-
It is recommended that you select the identity namespace that you believe has the most coverage in your event dataset.
-
-
In the Audience Analysis / Audience overview template (fixed to yesterday), use:
-
Current Members: who is still in the audience
-
Exited Audience: who left yesterday
-
-
Build your Workspace projects to:
-
Filter to profiles that exited Audience X yesterday, then look at:
-
Their behavior leading up to exit (last sessions, errors, price/offer exposure, channel mix).
-
Their post‑exit behavior (did they switch products, downgrade, go dormant).
-
-
Break that exited cohort down by region, device, tenure, value tier to find high‑impact pockets.
-
-
Use the insights for journey updates and win‑back audiences configurations in RTCDP or AJO.
Identity resolution considerations
“Who just left a key audience?”
“What did they do around exit?”
Exits measured on a stable login/CRM ID tend to reflect true behavioral change.
When someone leaves an audience on this ID, it usually means a real change (churn, downgrade, inactivity).
You can analyze their recent behavior to fine‑tune journeys and win‑back offers with confidence.
Exits are only visible where profiles and events share the configured ID and thus will require careful interpretation.
Use exited cohorts as a strong hint or signal, but it is recommended that you cross‑check with other data points before critical decisions.