Adobe Customer Journey Analytics has shipped three features that each solve a real, longstanding frustration. No fluff — here’s what they are, what they fix, and what you can do with them.
1. Audience Analysis: From Hypothesis to Proof
Experience League Link: https://experienceleague.adobe.com/en/docs/analytics-platform/using/cja-connections/audience-analysis/audience-analysis-overview
What it is
You can now ingest Adobe Experience Platform (RTCDP) audience segments directly into CJA as dimensions. Your audiences — the same ones you use for targeting and activation — show up as usable dimensions in Analysis Workspace, ready to slice and filter like any other data point.
What it solves
Until now, CJA and RTCDP lived in separate worlds. In RTCDP you’d define audiences and activate them for campaigns. In CJA you’d analyze behavior. But you could never close the loop: is this audience actually behaving the way we expect in our key journeys? Of course you could build the same audience with the attributes of the profile dataset in CJA, but you ended up with “duplicate” definitions.
How it works now
Once configured Audience Analytics, your AEP audiences become dimensions in Analysis Workspace. You can filter any analysis by audience membership, compare audiences side by side, and see how different segments move through the same journey.
How it works technically
When you configure Audience Analysis, CJA automatically adds two datasets to your connection. First, a lookup dataset that maps the segment membership key to a human-readable audience name — this is what makes your audience show up as a proper dimension label in Workspace rather than a raw ID. Second, the Profile snapshot dataset from AEP, which carries the actual audience membership data per profile.
The important detail for anyone watching their contract: because these are a lookup dataset and a profile dataset, they are not counted against your licensed row consumption. You get the audience dimension without paying extra rows for it.
Audience membership data is reprocessed nightly, so the data in CJA reflects audience state as of the previous day.
Use cases
Audience validation — You built a “High-Intent Shoppers” segment in RTCDP and activated it for a campaign. Now you can open CJA and actually see how that segment behaves: conversion rate, drop-off points, revenue per session. Not impressions. Actual on-site behavior.
Audience discovery — Instead of committing to one audience definition upfront, build an experimental dashboard with your key journeys and KPIs, then layer in different audience segments. Which one actually performs better in this journey? Let the data tell you, rather than your hypothesis.
Segment quality checks — Are users in your “Churned” segment actually behaving like churned users? Compare them against active users and see if your segment definition holds up in the real world.
The bigger shift here is conceptual: you move from “we think this audience fits this journey” to “we can see which audience fits this journey.”
2. Identity Stitching in the UI: Cross-Device Analysis Without the Support Ticket
Experience League Link: https://experienceleague.adobe.com/en/docs/analytics-platform/using/stitching/use-stitching-ui
What it is
You can now enable identity stitching — connecting anonymous and authenticated events across devices and channels into a single customer journey — directly in the CJA Connections UI. No more support tickets, no waiting.
What it solves
Identity stitching has always been one of the most valuable things CJA can do. A user browses anonymously on mobile, then logs in and converts on desktop — stitching lets you see that as one continuous journey instead of two unrelated sessions.
The problem was the process. Previously, enabling stitching required opening an Adobe support ticket, waiting for Adobe to set it up on the backend, and then ending up with two near-identical datasets in AEP — the original and the stitched version — which added cost and confusion.
How it works now
Inside the Connection settings, you enable stitching per dataset and choose your method: field-based stitching (using a persistent ID like a cookie and a person ID like a CRM ID) or graph-based stitching (using the AEP Identity Graph to resolve identities). That’s it. No ticket, no waiting, no duplicated datasets.
How it works technically
The following is based on a conversation I had with Adobe Product Managers — here’s a summary of how it works under the hood. Credits to Brian Au and Matt Thomas.
The controls haven’t changed — they’ve just moved into the UI and been standardized. You still choose your stitching method by selecting the Person ID: pick a dataset field or identityMap for field-based stitching (FBS), or Identity Graph for graph-based stitching (GBS). The replay window is also set there.
The most significant under-the-hood change is where stitching runs. Instead of producing a separate stitched dataset in AEP, stitching now runs directly in the CJA ingest path — same FBS/GBS engines as before, just operating inside the connection rather than outside it. This is why there’s no longer a duplicate dataset sitting in your AEP data lake.
Already using the “legacy” stitching method? You don’t need to migrate yourself. Adobe will migrate existing ticket-provisioned stitched datasets to the new UI-based experience on your behalf, likely later this year.
Use cases
Cross-device journey analysis — A user discovers your product on mobile, does research on desktop, and converts on tablet. Stitching connects all three sessions to one person. You see the full journey, not three fragments.
Web + call center linkage — Link anonymous web behavior (cookie-based) with authenticated call center records (CRM ID). Now you can see what someone was doing on your site before they called — and whether that influenced the call outcome.
Faster implementation — What previously took days through a support process now takes minutes in the UI. For teams managing multiple connections or iterating on data models, this is a meaningful time saver.
3. Multiple Dimensions in Freeform Tables: Simpler, Richer Tables (Limited Testing)
Experience League Link: https://experienceleague.adobe.com/en/docs/analytics-platform/using/cja-workspace/visualizations/freeform-table/freeform-table-multidimensions
This feature is currently in Limited Testing, so availability may vary by environment.
What it is
You can now add up to 5 dimension columns side by side in a single freeform table. Each row becomes a combination of all dimension values — for example, Channel × Device Type × Region — treated as a single data point.
What it solves
For years, the only way to analyze multiple dimensions together in a freeform table was through breakdowns: drill into a row, add a breakdown dimension, drill again. It works, but it creates deeply nested tables that are hard to read, hard to share, and painful to maintain.
Simple questions — “which channel, on which device, in which region converts best?” — required multiple breakdown layers or a separate panel. Neither was clean.
How it works now
Drag up to 5 dimensions into a freeform table, side by side. The table renders each combination as its own row. You can sort by any column (dimension or metric), apply filters, and even add breakdowns on top for deeper dives when needed.
Use cases
Campaign performance matrix — Put Campaign × Creative Variant × Device in one table. Sort by conversion rate. Instantly see which combinations work — no nested breakdowns, no extra panels.
Easier dashboards — Operational dashboards with multi-dimensional views used to require workarounds or complex panel setups. Now you build the table directly. Cleaner output, less maintenance.
Cohort-style analysis on the fly — Combine Acquisition Month × Product Category × Customer Tier in a single table to approximate cohort behavior without building an actual cohort. It’s not a full cohort analysis, but for quick explorations it’s significantly faster.
Summary
| Feature | The old pain | What changed |
|---|---|---|
| Audience Analysis | RTCDP audiences invisible in CJA | Audiences are now CJA dimensions |
| Identity Stitching UI | Support ticket + duplicate datasets | Configure stitching directly in Connections |
| Multi-Dimension Tables | Breakdowns only, nested and messy | Up to 5 dimensions side by side |
These aren’t headline features with flashy demos — they’re fixes to real friction points that CJA practitioners have lived with for a long time. If you work with CJA regularly, at least one of these probably made you think “finally.”