Workspace Intelligence and audiences
Give the AI features durable context about your product, competitors, target audiences, and research evidence.
Workspace Intelligence is your workspace's knowledge base: durable context that describes who you build for, what you're building, and what you've learned. AI-powered features — above all AI studies — draw on this context so their output reflects your actual product and users rather than generic assumptions.
It's available on team workspaces, for everyone except tester seats (Team Member seats can look but not edit). How many items of each kind you can keep active — and whether a type is available at all — depends on the plan; see Plan limits. The Overview page shows your most and least used intelligence and the evidence generated across the workspace.
The four item types
| Type | Describes | Details |
|---|---|---|
| Audiences | The user groups you build for — goals, pain points, behaviors, in structured sections | Audiences |
| Products | What you're building — description, links, documents, and notes | Products |
| Competitors | The alternatives — with typed research sources | Competitors |
| Evidence | What you've learned — findings linked to the items they're about | Evidence |
Attaching context to projects
Intelligence items are workspace-wide, but their day-to-day use is per project: attach the relevant audiences, products, competitors, and evidence to a project under Project Context and AI features working there use exactly that context (an AI study's Use project context option pulls it in, together with the project's goals and PRDs). Archived items can't be attached. Folders have a Folder Context too — projects created inside the folder afterwards inherit the folder's active items automatically; existing projects keep their own context.
Labels
Audiences and evidence items carry free-form labels for filtering and search. Labels are normalized — trimmed, de-duplicated case-insensitively — so B2B and b2b are one label.
Archiving
Archived items are hidden from pickers but keep their history; restore them at any time. Archiving frees the item's slot against the plan limit, which makes it the first move when a create is blocked — restoring is itself subject to the limit check.
In this section
- Audiences
Describe your user groups in structured sections — built by hand or extracted from uploaded persona documents — and power audience-mode AI studies.
- Products
Represent what you're building: description, links, uploaded documents, and notes that AI studies can draw on.
- Competitors
Track the alternatives your users compare you against, with typed research sources the AI can cite.
- Evidence
Store research findings as durable evidence items, linked to the audiences, products, and competitors they're about.
Related docs
- AI studies
Run AI-powered evaluations of your screens: audience studies simulate your target users, expert studies apply UX and accessibility review methods.
- Create and configure a study
Set up a new study inside a project, name it, and arrange its blocks in the study builder.
- Plan limits
How many projects, studies, audiences, competitors, evidence items, and products each plan allows — and what happens at the limit.