Martech Partners Contact
MartechPartners

Journey Optimizer November 2026: alert cards, transactional migration

Adobe Journey Optimizer · Editorial Team · 18 November 2026

The November 2026 Adobe Journey Optimizer notes put the most weight on three things: AI-prepared alert investigations on the home page, a wider migration path for transactional email, and richer inputs for Personalized optimization models. Almost every item is Limited Availability, so the practical question for most teams is which ones to ask Adobe for, and what has to be ready before the answer is yes.

Key takeaways

  • Journey Custom Action failure and Journey Anomaly Detected alerts can now produce an AI Recommendations card on the home page, with analysis already prepared in Coworker. It is Limited Availability, dated 16 November 2026.
  • Coworker gained three read-only orchestrated campaign skills on 3 November 2026. They help you look at campaigns, not change them.
  • The migration workspace now supports Campaigns, so transactional emails from another marketing platform can be converted into API-triggered campaigns. Limited Availability, 13 November 2026.
  • Personalized optimization models can take up to 50 profile attributes and up to 20 context attributes. Context works only for inbound-channel policies. Limited Availability, 3 November 2026.
  • Experience Manager Content Fragments accept attribute paths such as {{profile.person.name.firstName}} with no token map, but only when added directly to a message. Limited Availability, 5 November 2026.
  • Treat this as a request list for your Adobe representative plus a data and ownership checklist, not as a deployment project.

What the November 2026 Journey Optimizer updates change in practice

The release notes group the month into five areas: Journeys, Campaigns, Onboarding, Decisioning and Integrations. Read together, they say something about direction. Adobe keeps moving the work of investigating, auditing and explaining into Coworker, and keeps adding inputs to the real-time machinery underneath.

None of it rewrites how you build a journey. You will not open the canvas next week and find a different tool. What changes is the work around the build: what happens when a custom action fails at two in the morning, how you answer a stakeholder asking which campaigns are live on a channel, how long a platform move takes, and how much signal your decisioning models see.

That is also why the month is easy to over-read. Four of the five items are gated behind a conversation with Adobe. If you are not in the set of organizations that has access, nothing on your screen changes yet. Plan around the request, not the feature.

The notes also point back to October, when Coworker picked up skills such as Channel Builder, Business Performance Analysis, Journey Hygiene Analysis, Trace Journey Profile and Journey Execution Timeline. November extends the same pattern rather than starting a new one. If your team has not tried Coworker at all, start there before chasing the newest card or skill.

AI recommendation cards for journey alerts: who gets them and what to do

When a Journey Custom Action failure alert or a Journey Anomaly Detected alert fires, an AI Recommendations card now appears on the Journey Optimizer home page. Selecting it opens the analysis and suggested next steps that Coworker has already prepared, so nobody has to ask for an investigation first. You can then keep talking to Coworker to explore the findings. Availability date: 16 November 2026, and only for a set of organizations in Limited Availability.

The people who benefit most are the ones who get paged. Think of a team that sends a time-sensitive confirmation through a custom action to an external system. When that endpoint starts failing, the first twenty minutes usually go to working out whether the failure is on the Adobe side, the endpoint side or in the payload. A card that arrives with a first analysis shortens the start of that conversation. It does not remove the need for someone who understands your endpoint.

What to do before you ask for access:

  1. Confirm that your alerts are switched on and routed to people who will actually look at them. A smarter card on top of ignored alerts changes nothing.
  2. Write down who owns each custom action's downstream system. The recommendation can point at a likely cause, but a named human has to call the vendor or fix the endpoint.
  3. Decide who is allowed to act on a suggestion. Coworker prepares next steps. Whether a junior operator may pause a live journey on that advice is your policy question.
  4. Run one rehearsal in a sandbox where possible: break a test action on purpose and see what the team does with the output.

What not to do: do not treat the card as a diagnosis you can paste into an incident report. It is analysis to check. Verify against your own logs and the documentation Adobe links from the notes before you change a production journey.

Orchestrated campaign skills in Coworker: audits without API work

From 3 November 2026 you can find and inspect orchestrated campaigns by chatting with Coworker. Three skills cover it. Orchestrated Campaign List lists, filters and counts campaigns, for example by status or channel. Orchestrated Campaign Get returns one campaign's details, its lifecycle history and its execution history, and can show its activity graph. Orchestrated Campaign Workflow Renderer draws a workflow graph as a diagram.

All three are read-only. The notes say plainly that creating, editing or publishing campaigns through Coworker is not yet supported. Keep that in front of anyone who hears 'conversational' and imagines building campaigns by typing.

The value is live visibility without needing API access or clicking through the interface. A few examples of where that matters:

  • A channel lead asks how many campaigns are currently using a given channel. Instead of filtering screens by hand, someone asks Coworker to count them.
  • A compliance reviewer wants the lifecycle history of one campaign. The Get skill gives them that without a request to an engineer.
  • A new team member needs to understand how a campaign is wired. The rendered diagram is a faster starting point than a verbal walkthrough.

This is a good candidate for self-service by the internal team. It needs no integration work, and the risk is low because nothing can be changed. The one thing worth doing is agreeing on who may use it and what counts as an authoritative answer. If Coworker's count disagrees with a spreadsheet someone maintains, you want a rule for which wins.

Migrating transactional emails into API-triggered campaigns

The migration workspace now supports Campaigns. You can bring transactional emails you built in an existing marketing platform into Journey Optimizer and convert them into API-triggered campaigns. This is Limited Availability, with an availability date of 13 November 2026, and you need to contact your Adobe representative for access.

Transactional email is the part of a platform move that teams tend to leave until last, because it looks small and is anything but. Receipts, password resets, order and booking confirmations: each one is triggered by another system, usually has strict timing, and has a customer who notices instantly when it breaks. A migration tool that handles the conversion into API-triggered campaigns removes a chunk of manual rebuild.

It does not remove the need for judgment. The notes describe converting emails into API-triggered campaigns. They say nothing about validating content, sender setup or the calling systems, so assume that is still your job. Whatever the converter produces, the calling application has to start sending to a new trigger, and someone has to prove the output matches.

A sensible approach is to inventory first. List every transactional message, the system that fires it, its volume pattern and its business owner. Migrate the lowest-risk message first, run old and new in parallel where your design allows, and compare rendered output and delivery before cutting the sending system over. Keep the old path available until you have seen a full cycle of real traffic.

This is also where an implementation partner earns their fee on a platform move. The conversion is a tool job. The cutover plan, the testing across the sending applications and the rollback design are project work. If you are weighing that, the Journey Optimizer partners page is a reasonable place to start comparing who has done migrations like this.

Profile and context features for Personalized optimization models

For Personalized optimization, you can now enrich AI models with profile and context features, in addition to or instead of audiences. When you create a model you can pick up to 50 profile attributes from your profile schema, such as loyalty tier or churn propensity score, and up to 20 context attributes sent with the decision request, such as device type, language or page URL. The model uses those signals to score decision items in real time. Context features are only available for decisioning policies used in inbound channels. It is Limited Availability, dated 3 November 2026.

The practical change is that you are no longer limited to audience membership as the way to tell a model who someone is. An audience is a coarse yes or no. A loyalty tier or a propensity score carries more information, and a page URL or device type tells the model what the person is doing right now.

The catch is that the quality of the model's inputs becomes your problem. Before you select attributes:

  1. Check that the profile attributes you want are populated for enough of your profiles. A field that is empty for most people adds little.
  2. Make sure context attributes are actually sent with each decision request. If your web or app team does not pass page URL or language, the model cannot use them. That is an engineering ticket, and it needs scheduling.
  3. Confirm the channel. Context features only apply to policies used in inbound channels, so an email-only decisioning setup will not benefit from them.
  4. Be selective. The ceilings are 50 and 20, but a ceiling is not a target. Start with a handful of attributes you can explain to a colleague.

Measure it the way you would any decisioning change: hold out a comparison, give it enough traffic to settle, and look at the outcome your offers exist for. The notes do not state any performance uplift, so do not promise one internally.

Dynamic token substitution for Experience Manager fragments

In Integrations, you can now use attribute paths such as {{profile.person.name.firstName}} directly inside Experience Manager Content Fragments, with no token map needed. It works only for fragments added directly to a message, not for fragments served through Experience Decisioning. It is Limited Availability, dated 5 November 2026.

If your content team keeps copy in Experience Manager, this removes a mapping step that was easy to get wrong. The author writes the personalization where the copy lives and the message resolves it. That is a gain for teams that reuse the same fragment across many messages.

The exclusion matters more than the feature. If some of your fragments reach customers through Experience Decisioning, the same tokens will not behave the same way. Audit which route each important fragment takes before telling authors to rewrite anything. A worked example: a welcome-message fragment attached straight to an email is a fine fit, while a fragment returned as an offer through a decisioning policy is not covered by this note.

Agree one authoring convention and test with profiles that have missing names. Decide what your fallback copy says when a value is absent, because a greeting with a blank in it is worse than no greeting.

A rollout checklist and who owns each piece

Because most of this needs access first, the order of work is: request, prepare, test, then use. The table below pairs each change with the people it affects and the first sensible action.

Change Availability Who it affects First action Suggested owner
AI recommendation cards for journey alerts Limited Availability, 16 November 2026 On-call operators, journey owners Confirm alert routing, ask your Adobe representative about access Platform operations
Orchestrated campaign skills in Coworker Read-only, 3 November 2026 Channel leads, reviewers, new team members Agree who may use it and which answers are authoritative Marketing operations
Transactional email migration to API-triggered campaigns Limited Availability, 13 November 2026 Teams moving from another platform, application owners Inventory every transactional message and its sender system Migration lead with engineering
Profile and context features for Personalized optimization Limited Availability, 3 November 2026 Decisioning and personalization teams Check attribute quality, confirm context is sent on inbound channels Decisioning owner with web or app engineering
Dynamic tokens in Experience Manager fragments Limited Availability, 5 November 2026 Content authors, template owners Map which fragments are added directly and which go through Experience Decisioning Content operations

One email to your Adobe representative can cover all four gated items. Say which of them you actually have a use for, so the answer is about your situation and not a general inquiry.

What to ignore, and where a partner helps

Ignore any claim that these changes make Journey Optimizer 'autonomous'. The notes describe analysis, read-only inspection, a migration converter and extra model inputs. Each helps a person work faster. None of them replaces the person.

Ignore the Coworker campaign skills if you do not use orchestrated campaigns. Ignore the transactional migration if you are not moving platforms. Ignore the decisioning inputs until your basic decisioning is stable and your profile data is trustworthy. Features you cannot use well are noise.

Your own team can handle the access requests, the alert routing review, the Coworker permissions and the fragment audit. They need no outside help and would take a few days of focused effort.

A partner is worth the money where the risk is concentrated. That means transactional cutovers where several applications call new triggers, decisioning work where context attributes have to flow from front-end code, and any migration large enough to need a rollback plan. If you go that route, write the brief carefully. The Journey Optimizer RFP template gives you a structure for scope, systems and success measures, and the partner finder helps you shortlist. If you are unsure whether you need outside help at all, the advisor is a quick way to test that question.

The sensible stance this month is patience with a short to-do list. Ask for what you can use, fix the data and ownership gaps that would stop you using it, and leave the rest until your own roadmap needs it.

Sources

Frequently Asked Questions

Are the November 2026 Journey Optimizer features available to every organization?

No. The AI recommendation cards, the transactional email campaign migration, profile and context features for Personalized optimization models and dynamic token substitution for Experience Manager fragments are all marked Limited Availability. You need to contact your Adobe representative for access. The three read-only orchestrated campaign skills in Coworker are not marked that way in the notes.

Can Coworker create or publish orchestrated campaigns in Journey Optimizer?

Not yet. The three skills released on 3 November 2026 are read-only: they list and filter campaigns, retrieve details and history for one campaign, and render a workflow graph. Creating, editing or publishing campaigns through Coworker is stated as not yet supported.

How many profile and context attributes can a Personalized optimization model use?

When you create an AI model you can select up to 50 profile attributes from your profile schema and up to 20 context attributes sent with the decision request. Context features only work for decisioning policies used in inbound channels. The capability is Limited Availability.

Do dynamic tokens work for Experience Manager fragments served through Experience Decisioning?

No. The notes say token substitution with attribute paths works only for fragments added directly to a message. Fragments served through Experience Decisioning are excluded, so check how your fragments are delivered before planning around it.