Build better products with our product team
Update the Campaign Reporting UI to allow filtering by specific run timestamps rather than just aggregating metrics by day. Why is this feature important to you Enterprise orchestrations often run multiple times a day (e.g., every 6 hours). Without instance-level granularity in the UI, stakeholders face massive reporting blindspots and must rely on custom AEP query service backends to properly distinguish performance between multiple runs on the same day. How would you like the feature to work The Batch Instance selector in the reporting UI should permit filtering by precise execution timestamps, not just by the date. Current Behaviour If an orchestrated campaign runs multiple times a day, the OOTB UI only shows aggregated daily metrics because the Batch Instance selector filters by day, not by timestamp
Allow Orchestrated Campaigns to natively evaluate freshly updated table values mid-execution rather than relying on execution-time snapshots. Why is this feature important to you We often need to verify if a user achieved an action (like changing an order status or receiving a previous email) right before sending a subsequent message. Inserting wait periods of several hours just to natively re-query updated datasets severely degrades the real-time nature of orchestrated messaging. How would you like the feature to work Introduce a configuration that forces a live evaluation of specific linked relational datasets mid-canvas before proceeding to the next delivery node. Current Behaviour Because OC relies on a batch-oriented snapshot at execution time, it does not evaluate freshly updated table values mid-execution
Enable OCM to perform a live Unified Profile Service (UPS) lookup to check consent before every email send, rather than relying on a static batch snapshot. Why is this feature important to you Strict compliance with privacy regulations demands that if a user denies consent mid-execution, they are immediately suppressed. Relying on workarounds like injecting wait periods and rebuilding audiences before every single channel send creates incredibly bloated and inefficient workflows. How would you like the feature to work The engine should automatically verify current consent status via a live UPS ping immediately before executing any delivery node, identical to how an AJO Journey operates. Current Behaviour OCM operates on a batch-oriented model using a "snapshot" taken at publication or run time; if a user denies consent while a campaign is mid-execution, they will still receive the scheduled emails
Introduce a runtime profile status to natively block profile re-entry while a user is still actively in-flight within an Orchestrated Campaign. Why is this feature important to you For daily recurring campaigns, architects currently face overlapping, concurrent journeys because profiles are pulled in again before exiting the workflow. Resolving this currently requires maintaining a custom active flag explicitly in the data model and updating it via external systems, which adds unnecessary architectural overhead. How would you like the feature to work OCM should natively expose a runtime profile status (such as "active" or "in-progress") that acts as a built-in mechanism to natively block campaign re-entry while a user is still actively in-flight. Current Behaviour Unlike standard Journeys, OCM does not natively expose a runtime profile status, providing no built-in mechanism to natively block re-entry while a user is still in-flight
Ensure Enrichment activities only expose the specific relational row that matched the initial "Build Audience" criteria rather than bleeding into the entire related-record level. Why is this feature important to you Practitioners need precise downstream personalization. When the system qualifies a profile based on a specific condition (like an order status of "Pending") but exposes all orders associated with that user during enrichment, it necessitates building redundant filtering logic directly inside the Enrichment data setup to prevent personalization errors. How would you like the feature to work The Enrichment step should automatically inherit and restrict the relational data to only the specific row that matched the initial condition in the Build Audience step. Current Behaviour The "Build Audience" step only qualifies the Person ID, and the downstream enrichment step automatically exposes the broader related-record level for that person rather than restricting it to the matched row
Your ecommerce platform isn't slow because of traffic. It's slow because of architecture.Global ecommerce is heading toward $7.4 trillion by 2026. And the brands eating that market share all have one thing in common: they ditched monolithic platforms years ago.As someone who architects large-scale commerce platforms daily, Adobe Commerce, Shopify Plus, headless builds, here's what I'm seeing on the ground:The shift to Composable Commerce is no longer a trend. It's the baseline.What composable actually means in practice:→ Your frontend (Next.js / React) is completely decoupled from your backend→ Your commerce engine handles cart, catalog, pricing, nothing more→ AI and personalization run as a dedicated layer, not bolted on plugins→ ERP, CRM, PIM, OMS connect via event-driven APIs not fragile point-to-point integrationsAnd the new frontier nobody's talking about enough: Zero-Click Commerce.AI agents are starting to shop on behalf of consumers. They don't browse. They decide. If your product catalog isn't structured, API-accessible, and compatible with AI reasoning models, you won't even be in the consideration set.Your competitors aren't just competing for clicks anymore. They're competing to be the brand that AI recommends.65% of brands report higher conversion rates after implementing personalization. 67% of enterprise retailers are already on headless. The window to get ahead of this is closing.The stack you build in 2026 will determine your commerce revenue in 2028.
Description - (I am sure this has been asked many times, so I just wanted to revive the idea) We need an automatic "Previous Period" date range component that automatically creates a dynamic lookback window relative to the selected date range on the panel. Why is this feature important to you It is basically not practicable to have fixed previous date range components in place that only work until one changes the panel date range. Example Freeform table using the last 7 full days (Panel date range) vs the previous 7 days (custom date range in the workspace project). Now, someone decides to change the date range on the panel to "last 30 full days" Result the custom date range still displays only 7 days, the rest is 0 the comparability between panel date range and freeform table is completely useless this becomes even more obvious / erroneous if you are applying a summary change visualization on the two columns, not taking the same number of data points into consideration How would you like the feature to work The "Previous Period" date range automatically calculates the previous x days / weeks / months before the currently selected date range, based upon that value. Should I change the panel date range, this will automatically update the data that has been filtered by this date range e.g., in a "Summary change" visualization Current Behaviour once you change the panel date range, your data is messed up, becoming unusable this is extremely annoying for all levels of analytics consultants. I don't know how many times I have explained this to clients or had the same issues myself other analytics tools like Looker Studio do not seem to have that problem, see https://support.google.com/looker-studio/answer/9272806?hl=en For an analytics solution as mature as Analysis Workspace, this should be a standard feature. @ericmatisoff your two cents on this idea would be amazing. You seem to have the best connections to the product team
We should have the option to disable / remove the opt-out link in scheduled projects. We’re sending scheduled projects to distribution lists with 200+ recipients. But we can’t manually add 200+ recipients in the Adobe scheduler efficiently. Therefore we only have a single email address in the scheduler (an internal distribution list).Here’s the issue: everyone receives the same scheduled project containing the same opt-out link. If only one person of the 200 recipients clicks on the opt-out link the whole distribution list is removed from the scheduler which this is stopping the scheduled report for the whole list on almost a daily basis.
Now that we can add Brand guidelines to Workfront it sure would be great if we could generate reports that use the Brand colors. Maybe offer that as an option that you could toggle on/off when generating a report.
Similar to how it shows there is a document attached (1), I am wondering if the same feature can be added under updates. So instead of it being blank next to updates, it would automatically populate the number of updates (1), another update is added (2), and so on.
It will useful to have these features in Advanced BI Reporting1. Possibility to include custom data fields2. Email CTA text be added to a data field in the same way as Email Links
A huge improvement to book a meeting in Dynamic Chat/Brand Concierge would be for a native LeanData integration so all routing flows through the same routing tree instead of having to build out a new routing in Dynamic Chat which is extremely limited in routing rules. This integration is available on all other chat bot tools including Qualified, 1Mind and Drift. LeanData can also work on Marketo forms so adding the ability in Dynamic Chat would be great. This is a small improvement that would bring great value to organizations to ensure leads are routed correctly regardless of the channel the prospect interacts with.
It is a very chaotic job for a Workfront System Admin to perform clean up activity or to offboard a group and its objects.It solves a admin pain point, if there is a tool/functionality added to Adobe Workfront for cleaning up the old data such as Projects, users, objects which are deactivated or not used in the last 6 months or a year.
DescriptionThe executionMetadata helper function in AJO is extremely valuable for dynamically capturing and storing custom key-value pairs in the message execution context.In our organization, the primary use case is tracking user exposure to specific fragments and enabling downstream personalization logic without the need to create additional journey branches or duplicated nodes.However, currently the execution metadata is only populated in the Feedback dataset (Sent, Delivered, Bounce, Error, Exclusion). It is not populated in the Tracking dataset for interactionType events (Opens, Clicks, Spam, and Unsubscribe). Why is this feature important to youThis limitation significantly reduces the analytical value of execution metadata.Today, if organizations want complete reporting across both delivery and engagement metrics, the only viable option is to create separate journey nodes for each variation or fragment exposure scenario. This increases journey complexity, maintenance effort, and operational overhead.If execution metadata were also propagated to the Tracking dataset, organizations could:Build full-funnel CJA reports using metadata dimensions Analyze opens, clicks, unsubscribes, and spam events by fragment exposure Reduce the need for duplicated nodes and branching logic Enable more scalable personalization and experimentation strategiesThis would make execution metadata far more powerful as both an orchestration and analytics capability. How would you like the feature to work?Ideally, any key-value pair defined through executionMetadata should automatically persist across both the Feedback and Tracking datasets. Current behaviorCJA reports based on execution metadata return zero values for engagement metrics, making it impossible to build complete funnel analyses using metadata-driven logic. As a result, organizations are forced to create additional journey nodes to enable proper reporting.
Idea summary:Workfront Fusion should support non-client-secret authentication patterns for Azure DevOps integrations using Microsoft Entra ID, specifically OIDC-based federated authentication / workload identity federation.Problem:Today, the Fusion Azure DevOps Entra authentication flow depends on client secrets. In our environment, client-secret authentication is no longer permitted under enterprise security policy. Because of that, we cannot use the current Fusion connection method in a compliant way.Requested capability:Add support for secretless authentication so Fusion can obtain tokens from Microsoft Entra ID without relying on client secrets.Why this matters:- Our current integration path is blocked by security policy- This has been a long-running blocker for our Fusion to Azure DevOps use case- It affects business operations and slows our ability to expand Workfront usage- We need either native support for this auth pattern or a clearly supported alternativeDesired outcome:Support OIDC-based workload identity federation / federated credentials for Microsoft Entra ID in Fusion integrations where client-secret authentication is not allowed.Additional context:This request is tied to enterprise security requirements and is not just a convenience enhancement. A supported secretless authentication model is required for compliant implementation in our environment.
DescriptionWe have identified UX inconsistencies in the Tag Selector component within the new Content Fragment Editor in AEM as a Cloud Service.These issues are causing confusion for authors and may lead to incorrect tagging of content.We have identified UX inconsistencies in the Tag Selector component within the new Content Fragment Editor in AEM as a Cloud Service.These issues are causing confusion for authors and may lead to incorrect tagging of content. Issue 1: Tag Selection and Expansion BehaviorCurrent BehaviorClicking on a tag label performs two actions simultaneously: Expands the tag hierarchy Automatically selects/adds the tag value Clicking specifically on the left arrow icon only expands the tag (expected behavior)Expected BehaviorClicking on a tag label should only expand/collapse the hierarchy Tag selection should be a separate explicit action (e.g., checkbox selection)ImpactAuthors may unintentionally select incorrect tags Leads to data inconsistency in content tagging Increases manual correction effort and validation overheadIssue 2: Incorrect Expansion Indicator (Arrow Icon)Current BehaviorTags display an expand arrow icon, even when: No child tags are available Expected BehaviorExpand arrow should be displayed only for tags that have child nodesImpactCreates confusion for authors Leads to unnecessary clicks and poor user experience Comparison with Classic EditorClassic CF Editor provides: Clear separation between expand vs select actions Checkbox-based selection, making intent explicit More intuitive and predictable UI behavior Steps to ReproduceOpen Content Fragment in the new CF Editor Navigate to the Tag Selector Click on a tag label Observe that the tag is both selected and expanded Expand tags that do not have children Observe that an arrow icon is still displayedProposed Enhancement / RecommendationDecouple tag expansion and selection behavior Introduce explicit selection mechanism (checkbox or similar) Remove expansion arrows for tags without children Align UX with classic CF editor behavior
I have used the experiment feature quiet a lot for the e-mail channel in our different onboarding journey’s (batch audience, sending daily) and recurring sales journeys. However, I have to stop my journeys and duplicated them and then adjust them each time I want to scale a winner, as the great scale the winner feature is not supporting outbound marketing e-mails yet (only api-triggered). There is a huge potential in getting the feature for this type of channel, as I believe this must still be the largest channel for majority of companies? I really hope this can be considered for the roadmap. I’ve eagerly checked the release notes each month since we got to be beta testers of Experimentation Accelerator about a year ago. Bonus: I’ve also just launch a test of the new Experiment feature inside the new Optimize node, to test different journey paths. My test is live now, but also doesn’t look like there a scale the winner feature for this one?Thanks for all the work on experimentations non the less. This is the fun part of the job :-)//Jonathan
I have created a custom field called "Product Scope," which is a multi-select dropdown. I want to track the changes made to this field, so I added it to the Update Feeds.When I update the field, I can see the logs in the Updates section of a project. However, when I remove 2 values and select 2 different values, it creates 4 separate log entries.Is this the expected behavior? The audit log says, "User updated Product Scope to Value1," which implies that the Product Scope is now only "Value1." However, since this is a multi-select field, there are other selected options as well.Do you also see this in the Updates section for Multi-select fields?
Several reports, particularly the correlations, subrelations and complex filters take a good amount of time to fetch. The user ends up navigating to some other work / different site to use the waiting time instead of seeing the waiting icon, sometimes leading to a session expiry. It would be much more useful if the user could be educated and entertained that time in that time by showing up some DidYouKnow stuff about Omniture, some Omniture/SiteCatalyst IQ Puzzles, or simple small games like TextTwister, Tiles etc which could be paused and kept to side as soon as the report arrives. (Omniture - You could even use it to upsell Discover, T&T etc.) The Exit rates from Omniture would go down significantly if that is done.
Currently [1] the option “Restore content” in the Cloud Manager gives you the opportunity to restore content of the whole environment. Here the problem is that you can only restore ALL content. If you are hosting just a small website, this would be not a problem. But if you imagine an enterprise setup for a company with several business groups who act independently, having multi-language and multi-sites, this option is not feasible. Imagine the following: you have a business group under the structure /content/business-a - here some content accident occurs which needs a content restore. Now, the content restore would effect everything under /content which means that you have to inform and coordinate ALL businesses about the restore as the work after the date where issue occurred and the point were you start restore will be gone and needs to be redone. Additionally, the editors a spread all over the world, so you can not really be sure that you reach everybody to inform them in time. So, what is missing is an option to restore content from a specific path only - just matching the content on which a problem occurred. We having that you could really just restore the effect path(s) and not anything else.[1]https://experienceleague.adobe.com/en/docs/experience-manager-cloud-service/content/operations/restore
We'd like the ability to add external URLs (hyperlinks) as items within the Documents section of projects, tasks, and issues — similar to how OneDrive allows linked files alongside uploaded documents.Use Case:Our teams frequently reference external resources (SharePoint pages, OneDrive folders, briefs hosted outside Workfront, vendor links, etc.) alongside project work. Right now there's no native way to surface those links in context — users either have to add them to custom form fields or drop them in the Updates section, both of which reduce discoverability.Expected Behavior:In the Documents section: a user can add a URL that appears as a linked item alongside uploaded files, with a display name and the external linkBusiness Impact:Reduces context-switching and improves discoverability of related resources. Teams are currently using workarounds (custom fields, pinned Updates comments) that are inconsistent and easy to miss. Sharepoint option for adding links as a line item in the Document LibraryWorkfront has many options but missing the ability to add links
Agentic AI is rapidly transforming marketing by enabling autonomous decisioning, content generation, audience creation, and journey orchestration at unprecedented speed. Within Adobe Experience Platform (AEP), Adobe Journey Optimizer (AJO), and Real-Time CDP, these capabilities are becoming increasingly real—not theoretical.However, with this acceleration comes a critical challenge: control, trust, and governance have not evolved at the same pace as AI capabilities.Without the right governance framework, agentic AI can introduce unintended risks across:Brand safety and consistency Regulatory and compliance adherence Customer experience integrity Decision transparency and explainabilityThe key question is no longer “Can we use AI?” but rather:“How do we scale AI responsibly while maintaining control, trust, and accountability?”This article outlines practical governance models that enable organizations to scale agentic marketing safely and effectively.Why Governance Becomes Critical in Agentic MarketingTraditional marketing workflows were built around:Human approvals Linear campaign execution Static segmentation Pre-defined journeysAgentic AI introduces a fundamentally different paradigm:Autonomous agents making decisions Real-time audience evolution Dynamic journey orchestration Continuous optimization loopsThis shift removes friction—but also removes traditional control points.Without governance, organizations risk:AI-generated content deviating from brand tone Audience misuse or unintended targeting Regulatory violations in real-time decisioning Lack of visibility into “why” an AI made a decisionGovernance is no longer a checkpoint.It becomes an embedded system design principle. From Approval Workflows to AI GuardrailsTraditional marketing governance relies heavily on approval workflows:Content review cycles Campaign sign-offs Manual QA processesWhile effective in deterministic environments, they do not scale in agentic systems.In AI-driven marketing, governance must evolve into guardrails, such as:Policy-driven decision constraints Real-time content validation Identity and audience usage rules AI model usage boundaries Automated compliance enforcementInstead of asking:“Did someone approve this?”We must ask:“Did the system operate within defined safe boundaries?”This shift enables speed and safety simultaneously.Human-in-the-Loop vs Human-on-the-LoopA key distinction in agentic governance is the operating model for human involvement.Human-in-the-Loop (HITL)Humans actively participate in every decision cycle:Review before execution Approval-based workflows Manual intervention requiredBest suited for:High-risk regulated decisions Sensitive customer communications Early-stage AI adoptionHuman-on-the-Loop (HOTL)Humans define constraints, monitor outcomes, and intervene when needed:AI operates autonomously within guardrails Humans supervise system behavior Exception-based intervention modelBest suited for:Real-time personalization Journey optimization Audience activation at scaleThe Future: Hybrid Governance ModelThe most scalable model is a hybrid approach:HITL for strategy and policy definition HOTL for execution and optimization AI agents operating within governed boundariesThis is where true agentic marketing becomes viable at enterprise scale.A Practical Governance Framework for Agentic AI in MarketingA scalable governance model for AI-driven marketing typically includes four layers:1. Data Governance LayerIdentity resolution rules Consent and privacy enforcement Data quality validation Audience eligibility constraints2. AI Governance LayerApproved model registry (LLMs, decision models) Prompt and output constraints Model usage policies (what AI can/cannot do) Explainability requirements3. Experience Governance LayerBrand tone and content guidelines Journey logic constraints Channel-specific rules Personalization boundaries4. Operational Governance LayerHuman oversight mechanisms Audit logs and traceability Exception handling workflows KPI monitoring and drift detectionScaling AI Responsibly Across Marketing OrganizationsOrganizations that successfully scale agentic AI do three things well:They embed governance into architecture, not processes They shift from manual approvals to automated guardrails They define clear boundaries for autonomous AI behaviorThe goal is not to slow AI down—but to make it safe to scale it faster.Final ThoughtAgentic AI represents a fundamental shift in how marketing operates. But without governance, speed becomes risk.The winning organizations will not be those who adopt AI the fastest—but those who design systems where AI can operate safely, transparently, and responsibly at scale.Governance is no longer a constraint on innovation.It is the foundation that makes innovation sustainable. I’d be interested in hearing how others in the community are approaching governance for agentic AI in marketing. What frameworks or guardrails are you putting in place to balance speed, control, and trust?
Small automation win I wanted to share with the community We recently set up an automated notification in Workfront that triggers the moment a proof is approved — instantly alerting both the content owner and the project owner.Sounds simple, right? But before this, those two people were either:— Waiting on manual follow-ups— Finding out late and losing time on next steps— Relying on someone to remember to ping themNow the moment a proof is approved, the right people know. No chasing, no lag.Sometimes the highest-ROI automations aren't the complex ones — they're the ones that eliminate a small, repetitive friction point that was quietly slowing everything down.Anyone else has quick wins like this? Would love to hear what small automations have made a big difference on your team. 👇
In the Marketo Admin panel under Integration > Salesforce > Edit OAuth Consumer Info, both the Consumer Key and Consumer Secret fields are displayed in plain text. This is a security concern — OAuth secrets should be masked by default (e.g., ••••••••••) with an optional "Show" toggle, consistent with how most platforms handle sensitive credentials. Current Behavior: - Navigate to Admin > Integration > Salesforce - Click "Edit OAuth Consumer Info" - Both Consumer Key and Consumer Secret values are fully visible in plain text Expected Behavior: - Consumer Secret should use a password-type input field, masked by default - Consumer Key should ideally also be masked - An optional "Show/Hide" toggle would allow authorized admins to reveal the value when needed Why this matters: - Screen sharing, screenshots, and over-the-shoulder exposure can inadvertently leak credentials - Security audits and compliance reviews (SOC 2, ISO 27001) flag plain-text credential display as a finding - Other Adobe products and most SaaS platforms already mask API secrets in their admin UIs - This is a low-effort, high-impact UX improvement that aligns with OWASP secure design principles Workaround: None — there is no way to mask these fields today. Reference: https://experienceleague.adobe.com/en/docs/marketo/using/product-docs/crm-sync/salesforce-sync/log-in-using-oauth-2-0
Document Proofs do not currently “honor” the User Settings around notifications. This is an issue for users who need broader visibility to Document Proofs, but do not require detailed summaries. Is this something Adobe Workfront team can look into resolving? This resolution would help prevent unnecessary email disruption for our management/leadership team that need visibility to Document Proofs without the detailed email summaries.More detail on case study: We have Workfront contributors that will view a monthly report that contains a proof link, where they can view a document proof. For these users, all User notification settings set to off. However, when they click to view the Document Proof, Workfront is adding them as a “Reviewer” to the Document Proof and automatically subscribing them to receive Daily Summaries for that project. We tested resolution by keeping them as Reviewer with Proof notification settings disabled. [However note: this resolution would be a challenge, as Document Proofs do not recognize Group settings, and would require us to setup each user as Reviewer with Disabled notifications for every project. This would be incredibly cumbersome for every project] This is an issue for our process, as these particular users require unnecessary email disruption and do not need visibility to all the detailed comments in the Document proofing process / daily summary.
Already have an account? Login
No account yet? Create an account
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.
