Adobe Analytics component descriptions and October 2026 dates

Adobe Analytics will be able to write descriptions for your dimensions, metrics, calculated metrics, segments and date ranges, with general availability on 28 October 2026. Alongside it, the October 2026 notes bring new MCP server permissions, an Adobe Brand Visibility integration and a delayed Coworker Chat. The descriptions feature is the one most teams will feel day to day, because it touches every Workspace user who has ever stared at a list of forty similarly named metrics.

Key takeaways

• Automatic component descriptions reach general availability on 28 October 2026; use the weeks before to clean up component names and ownership.
• The MCP permission item MCP Access is now MCP Full Access, and a new MCP Read Only item went generally available on 6 October 2026. Audit who holds which.
• Coworker Chat access to Analytics data began rolling out on 2 October 2026, but its general availability date is TBD, so don't schedule training around it.
• Adobe Brand Visibility is listed for October 2026 with no specific day and no documentation yet. Treat it as a pilot candidate, not a commitment.
• The Analytics 1.4 API and WSSE authentication were shut down on 31 August 2026, and legacy Report Builder was retired in June 2026. Confirm nothing in your estate still depends on either.

What does the October 2026 Adobe Analytics release actually change?

Four items are new or changed, and they land on different dates. Component descriptions are due on 28 October 2026. The MCP permission split was generally available on 6 October 2026. Adobe Brand Visibility is listed for October 2026 without a day. Coworker Chat began rolling out on 2 October 2026 with no general availability date.

The notes were last updated on 7 October 2026, and Adobe says they're revised several times a month under a continuous delivery model. Treat any date here as the state of the page on that day. Re-read it before you commit a project plan to anything.

The fix list is long but unglamorous. Classifications has the biggest block, with 22 ticket numbers. Report Builder has 8, Analysis Workspace 4, and Data Feeds and Data Warehouse 4. If you run heavy classification uploads or scheduled Report Builder workbooks, skim the ticket numbers against any open cases you've raised with Adobe. If you don't, there's nothing to act on.

Will generated component descriptions fix a messy Workspace?

They will help, but they won't fix it. A description answers the question an analyst asks when they open a component picker: what is this, and should I use it? Adobe's notes frame it exactly that way, aimed at organisations with large component libraries. A generated sentence is a better starting point than an empty field, which is what most calculated metrics carry today.

What it can't do is decide which of your fourteen variants of 'Orders' is the right one. A tool writing a description can only work from what exists. If a calculated metric is called 'Revenue v2 FINAL (use this)', the description won't rescue it. The naming and ownership work is still yours.

Here's a worked example. A retail team has three segments that all contain 'mobile' in the name: one built on device type, one on a campaign parameter, one on app-only traffic. Analysts have been picking one at random for two years. Generated descriptions might make the differences visible. They could just as easily describe all three in similar language, so someone who knows the definitions still has to read them.

Two practical questions the documentation should answer when it's published: whether generated text can be edited before or after it's saved, and who is allowed to trigger bulk generation. The notes don't say, and the link to documentation was still to follow. Check Adobe's page before you set policy.

Which dates in this release are firm, and which are not?

Only some of them. Keep rollout start and general availability separate, and don't treat a vague month as a date.

The Coworker Chat row is the useful lesson. A date that has already slipped once and is now TBD is a reason to wait, not a reason to book a training session. The streaming media row is the same story at a longer distance: the rollout began on 29 October 2025 and general availability is still undecided.

Are you really finished with the 1.4 API and legacy Report Builder?

Probably, but check. Both retirements are in the past now, which is when quiet breakage tends to show up. The notes say the Adobe Analytics API (version 1.4) and WSSE authentication reached end of life and were shut down on 31 August 2026, and that integrations built on them no longer function. The legacy Report Builder add-in was retired in June 2026.

The sort of thing that gets missed is the script nobody owns. A nightly export written years ago by someone who left, feeding a finance spreadsheet or a data warehouse load. If it stopped working in the last six weeks and nobody complained, either nobody reads it or someone is working around it by hand.

A short sweep is worth doing:

1. List every scheduled job, BI connector and workbook that pulls from Analytics.
2. For each, note which API or add-in it uses and who owns it.
3. Move anything on the 1.4 API to the Adobe Analytics 2.0 API, and anything on WSSE to OAuth-based authentication through the Adobe Developer Console.
4. Convert remaining legacy workbooks with the conversion feature in the new Report Builder, which the notes say is only available as an add-in through the Microsoft Store.
5. Record what you found, even if the answer is nothing.

That Microsoft Store detail matters in locked-down companies. The notes warn that many organisations need an internal approval process before the add-in can be made available. If your IT team hasn't approved it yet, that approval is now the critical path.

What should admins do about the MCP Read Only and MCP Full Access permissions?

Check who holds MCP Full Access, because every holder of the old MCP Access item now has it under the new name. The notes say the existing item was renamed and that users keep access to all tools, including those that create, change or delete components. The new MCP Read Only item gives access to all read-only tools and doesn't let a user create projects, segments or calculated metrics.

The sensible default is the narrower one. Most people who connect an assistant to the Adobe Analytics MCP server want to ask questions, not change a component library that dozens of colleagues rely on. Give those people MCP Read Only and keep MCP Full Access for a small group who build and maintain shared components.

This is a ten-minute admin task for most teams:

• Export or list the users with MCP Access today.
• Decide, per person, whether they create components or only read.
• Move readers to MCP Read Only and leave builders on MCP Full Access.
• Note the change in whatever access log your security team uses.

Don't over-engineer it. Nothing here requires a project, and there's no reason to bring in outside help for it.

Should you pilot Coworker Chat and Adobe Brand Visibility now?

Pilot Coworker Chat only if you already have access, and keep expectations small. The notes say Adobe CX Enterprise Coworker Chat can now perform advanced data analysis that was previously possible only in Analysis Workspace, drawing on your report suites and answering natural-language prompts. Rollout started on 2 October 2026, general availability is TBD, and the documentation link was still to follow.

The risk with natural-language answers isn't the technology, it's trust. If an answer disagrees with a Workspace panel, someone has to know which one is right. So test it against reports you already trust: pick five questions your team answers every week, run them in both places, and compare. Where the numbers differ, find out why before anyone uses the chat answer in a meeting.

Adobe Brand Visibility is a different kind of item. The notes say it connects with your Analytics data so you can measure how AI-driven discovery translates into website engagement and business outcomes. It's listed for October 2026 with no day, and documentation was still to follow. That's too little to plan around. A reasonable step is to nominate one person to read the documentation when it appears and decide whether the data you'd need is actually being collected.

What to ignore for now: any internal pressure to announce either one. A feature you can't yet describe the limits of isn't ready for a roadmap slide.

How should you roll this out and measure whether it worked?

Sequence the work around 28 October 2026, since that's the one date that changes everyday behaviour. A plain schedule works well.

Measurement doesn't need new tooling. Count the share of your top fifty most-used components that have a description a stranger would understand. Count the questions that analysts post in team chat asking which metric to use, before and after. Watch the number of duplicate components created per month. If the third number doesn't fall, descriptions aren't the lever; ownership is.

Don't bulk-generate descriptions for every component on day one and call it done. A thousand unreviewed sentences give a false sense of order. Start with the components that appear in dashboards executives actually open.

Where does an implementation partner help, and where can you do it yourself?

Most of this you can do yourself. Permission changes, a naming clean-up and a dependency sweep are admin work that a competent Analytics owner can finish in a few weeks. Hiring a partner to rename segments would be money badly spent.

A partner earns their fee where the work is wider than the product. If the dependency sweep turns up a dozen undocumented integrations, rebuilding them on the 2.0 API with proper OAuth setup is real engineering, and a team with limited developer time may want help. The same goes for a governance model across several report suites and business units, where the hard part is agreeing on definitions, not clicking buttons.

If you do go looking, write down the outcome first: which integrations, which report suites, which governance gaps. Our Adobe Analytics RFP template gives you a structure for that brief, and the Adobe Analytics partners page is a place to start comparing firms. If you'd rather narrow the field by region or size, find partners lets you filter.

Whatever you decide, keep one rule: the people who understand what a metric means should approve its description. Automation can draft the sentence. It can't know your business.

Sources

• Adobe Experience League: Current Adobe Analytics release notes — October 2026 release notes, last updated 7 October 2026: component descriptions, MCP permissions, Brand Visibility, Coworker Chat, end-of-life notices.