The UserVoice agent (Beta) lets you ask natural-language questions about your data — ideas, accounts, users, and themes — and get answers grounded in your actual UserVoice data.
Instead of manually filtering idea lists or exporting to a spreadsheet, you can ask:
- "What are our top 10 requests from enterprise accounts by revenue?"
- "What is Acme Corp asking for that we haven't responded to?"
- "Which themes have the most supporter activity this quarter?"
- "What has Sarah at Acme been asking for?"
The UserVoice agent reads your data in real time, reasons over it, and responds with prose summaries and clickable tables — all in a persistent, multi-turn chat interface.
What you can access
| Area | What the UserVoice agent can access | Key questions it answers |
| Ideas | Your full feedback backlog — titles, descriptions, comments, internal notes, statuses, dates, labels, themes, vote counts, and revenue weights | Top requests by votes or revenue; ideas without a response; recently changed statuses; trending ideas; ideas matching a keyword |
| Accounts | Your customer accounts — revenue, lifecycle status, supported idea counts, last activity, CRM data (plan, segment, industry), and your custom account fields | Top accounts by revenue; what a specific customer wants; accounts with the most open requests; churned accounts that wanted a feature |
| Users | Individual users — both end-users (idea creators and supporters) and admins — with name, email, linked account, and idea activity | Who submitted or voted on an idea; what a specific person has asked for; most engaged contributors at a given account |
What it can do:
You don't need to know how the agent works under the hood — just ask in plain language and it figures out which data to pull. This section describes what's possible and the limits that apply, so you know what to expect.
Searching your feedback (ideas)
Ask for lists of ideas filtered by status, category, forum, label, idea list, account, keyword, or date range, sorted by votes, revenue, recency, activity, comment volume, what's trending, or least recent activity.
- "Top 20 open ideas by votes in the Integrations category."
- "Unanswered ideas from enterprise accounts, sorted by revenue."
- "Ideas marked Planned in the last 30 days."
Searches open ideas by default (published, approved, and under-review) — ask explicitly to include closed or merged ideas.
For "show me all…" requests the agent pages through results, up to 200 ideas per answer. To count ideas that mention a term ("how many ideas mention SSO?"), it counts only ideas that contain that term.
Counting and breakdowns
Ask "how many ideas did we get by month?" or "break down open ideas by category" and the agent returns a breakdown table — by month, week, or year, status, category, forum, or source — scoped by the same filters as a search.
Reading an idea in depth
Ask a "why" question and the agent reads an idea's full description, customer comments, and internal admin notes to surface the real pain points — not just the title. It also reads the customer feedback records attached to the idea. Ask it to read everything and it reads up to 100 comments, notes, or feedback items in full, and tells you how many it read ("read 66 of 66 comments").
- "What are customers saying about the top API idea?"
- "Why do customers want this feature?"
Finding accounts
Filter your customer accounts by revenue, lifecycle status, engagement, CRM data, or your own custom account fields, and sort by revenue, LTV, recency, last login, or name. Typing the first few letters of a name finds the account. Custom fields work with text, number, and date values, including relative dates like "in the last 30 days" or "in the next 90 days."
- "Top 10 active accounts by revenue."
- "Show me churned accounts that wanted Salesforce integration."
- "Which enterprise-tier accounts renewing in the next 90 days support ideas we haven't planned?"
Defaults to active accounts (ask for churned, prospect, or lost). Account answers show when the account last supported an idea and, with Gainsight, how recent its health data is. Revenue and LTV fields require the revenue permission; health signals require a Gainsight integration.
Finding people (users)
Look up a person by name, email, linked account, or the ideas they've interacted with — useful for resolving "who is Sarah?" before drilling in.
- "Who is John Smith?"
- "Which users at Acme have voted the most?"
By default it includes anyone who engaged — voted or commented; ask for contributors to include idea creators who haven't voted or commented. A deep-dive on one person shows up to 10 ideas they created and 10 they support (open ideas).
Working with labels
Reference a label by name in your question and the agent resolves it automatically — matching is case-insensitive against the label name.
Detecting themes
Ask an analytical question and the agent clusters the relevant ideas into pain-based themes with impact and a representative quote. See How the agent detects themes below.
Making changes
Ask the agent to make a change and it proposes the change and shows exactly what will happen before anything is touched.
- Capture feedback: create a new idea (as yourself or for a customer), add a person's words to an existing idea, add voters, or flag ideas as important and remove flags.
- Organize ideas: save ideas as a new list, add ideas to an existing list or remove them, and update labels.
- Triage: set the public status your customers see or the internal status (with the status's default message, your own message, or no message), assign an owner, or edit one idea's title, description, or category. Setting a closing public status closes the idea, and any other public status reopens it.
- Link to engineering: link an idea to an existing Jira issue or Azure DevOps work item by its key, ID, or URL.
- Clean up: merge duplicate ideas into one, delete ideas, or mark ideas as spam.
Before it creates an idea, the agent checks for an existing one about the same request and shows similar ideas on the card. You can name someone who isn't in UserVoice yet by email, and the agent adds them — the card tells you first.
"Label these 'needs-scoping' and save them as a list called Q3 candidates."
"Set the internal status on these ideas to Under Review."
"Add this quote from Dana at Acme to the SSO idea."
"Flag these three ideas as important for Globex — they came up in our renewal call."
"Merge these two duplicates into the older SSO idea."
"Link the SSO idea to PROJ-123."
Nothing changes until you confirm. Requires feedback write access — read-only admins aren't offered these actions. Limits per change: 200 ideas for labels and owner; 25 ideas when setting a public or internal status, flagging, deleting, or marking spam; 50 ideas added or removed per list update; 25 people when adding voters; up to 10 duplicates merged into one idea; and one idea per Jira or Azure DevOps link.
Deleted ideas and ideas marked as spam leave your portal and lists, but their supporters, comments, and feedback are kept. The agent can't undo these changes, but you can undo them from each idea's page. When you merge, the duplicates' supporters, votes, comments, and feedback move to the idea you keep.
About email: changes send the same notifications as doing them by hand. New ideas in public forums go live right away and notify admins who get new-idea emails. A public status change posts its message on your portal and emails the idea's contributors; supporters are emailed only if you ask, and the card tells you who will be emailed. An internal status change emails your team only if you ask. Status changes can't be undone, so check the card before you confirm.
About linking: linking to Jira updates the issue's UserVoice links field. Linking to Azure DevOps adds a UserVoice tag, a link to the idea, and a comment with your name and the idea's title and description to the work item.
About spam: marking an idea as spam sends its text and its author's email and IP address to our spam-filtering service (Akismet) to improve spam detection. The confirmation card says this before you confirm.
Idea tables
When your question calls for a list of ideas, people, or accounts, the agent renders an interactive, clickable table inline. See Understanding what you see below.
Opening the agent from your ideas grid
Open the agent from the ideas grid and it already knows which ideas you're viewing or have selected. Ask "what do these have in common?" or "which of these matter most to our biggest accounts?" without copying anything over.
How the UserVoice agent detects themes
When you ask an analytical question, the agent reads the actual ideas in scope — their titles, bodies, and customer comments — and analyzes them on the spot to find recurring pain points. It's not pulling from a pre-built report; the analysis is generated fresh each time you ask.
The work is split in two:
- The agent clusters and names. It groups related ideas into themes, writes a specific pain statement for each, judges its intensity, and pulls a real, verbatim quote from a customer comment as evidence.
- The agent does the math. Every number you see — supporters, accounts, revenue, dates, admin activity — is calculated directly from your database, not estimated by the AI.
Two things worth knowing: a theme only forms when at least two genuinely related ideas support it, and ideas that don't fit a clear pattern are intentionally left uncategorized rather than forced into a theme. Because the analysis is generated live, results can vary slightly between runs. Each idea belongs to at most one theme. For large sets, the agent analyzes the 100 most-supported ideas and says so, and shows live progress while it works.
Click a theme in the table to see the ideas behind it, or pick a lens to go deeper: the workarounds customers use today, devil's-advocate challenges, research questions, the value of solving it, or sub-themes.
How these differ from your Themes & Trends report
The agent's themes are generated live, scoped to the specific question you just asked, and shaped to drive a decision — so instead of a category label you get a pointed pain statement, enriched with impact (supporters, accounts, revenue, admin activity) and a representative quote.
Differences are expected: the agent analyzes the slice of ideas relevant to your question at the moment you ask, using its own method and a more specific naming style, rather than reading back your saved report. The agent's themes are also ephemeral — they aren't saved and won't appear in Themes & Trends, and your Themes & Trends reports won't change because of them. Your Themes & Trends data stays exactly as it is.
Understanding what you see
Streaming responses
The agent shows its reasoning in real time. Short narration appears between steps — e.g., "Let me check the Onboarding category..." — followed by the answer. This lets you catch cases where the agent interpreted your question differently than you intended.
Inline idea tables
When your question calls for a list of ideas, the agent renders a clickable table inline. Columns are chosen to match your question — a revenue query shows Revenue; a staleness question shows last activity date. Each row links directly to the idea in UserVoice. Right-click any idea the agent mentions to Open idea sheet, Open in new tab, or Copy admin link.
Tables always show live data. The agent saves which ideas were in the table and re-pulls their current details (votes, status, owner, etc.) each time you reopen a conversation. The written analysis, by contrast, is saved exactly as the agent wrote it. So the table reflects your data as it is now, while the commentary reflects it as it was then — treat the table as the source of truth for current numbers.
Save as List: With nothing checked, the button reads Save all N as List and saves the whole table. Check rows to save just those. You'll be asked to name the list, and a "Saved as '…'" link appears inline — it doesn't navigate you away. The button is hidden for admins without feedback write access. The list is saved to Idea Lists. Tables show 10 rows per page; Copy table, Download CSV, and Save as List always act on every row.
Confirming a change
When the agent proposes a change, it shows a confirmation card: a title describing exactly what will happen (e.g., "Set internal status of 14 ideas to 'Under Review'"), the ideas affected (up to 10 listed, with a "+N more" tail), and for a single-idea edit, a before → after diff. When the agent creates an idea, the card also lists similar existing ideas so you can avoid a duplicate, and it warns you if it will add a new person to UserVoice. The card reads "Awaiting your confirmation" until you act. The confirm button names the action (for example Update labels, Create idea, or Flag as important) and turns red when the change removes something. Cancel drops the proposal and changes nothing.
Once you confirm, the card turns into a receipt in place: status flips to Completed, followed by a one-line summary, any side effects (such as a new user it created), and the ideas that were touched — each one clickable, opening the idea directly.
Rating and copying answers
Rate any answer with 
(Good response) or 
(Bad response), and optionally add details about your rating. This tells us whether that specific answer was helpful or wrong. Use Copy message to copy an answer with its formatting intact, or Copy table to copy a table.
What the UserVoice agent cannot do
Beyond the changes above, the agent cannot (yet):
- Split an idea into several ideas
- Edit existing users or accounts (it can add a new person by email when you capture their feedback)
- Create, edit, or unlink Jira issues or Azure DevOps work items (it can link an existing one)
- Send a message to anyone (beyond the notifications a change sends, such as notifying supporters about a merge you asked for), or escalate to a human
- Undo a change it made (undo deletes, spam, and merges from the idea's page instead; status changes can't be undone)
Ask for any of these and it says plainly that it can't, in one sentence, and stops — it won't invent a workaround or tell you how to do it in the UI.
The agent only answers questions about your UserVoice data. It won't function as a general-purpose AI assistant.
Tips for better results
Start from a starter prompt. Not sure what to ask? Open a new chat and pick a ready-made prompt from one of the categories: Get started, Revenue impact, Triage, Account insights, or Prioritize. When the agent is open next to your idea grid, Edit selection offers prompts for the ideas you've selected. Fill in the blanks (a feature or topic, a date range, a status, a label, an account, or an owner) and send. Revenue prompts note when your workspace has no CRM revenue data.
Be specific. "Show me top ideas by revenue" returns better results than "what should I work on?" Give the agent a direction to search in.
Use your own terminology. The agent knows your category names, forum names, label names, and the contents of your user feedback. Say "ideas in Onboarding" or "the Mobile theme" — it resolves them automatically.
Ask follow-ups. You don't need to restate context. After asking about Acme Corp, you can follow up with "what about ideas with no response?" and the agent stays on topic.
Respond if the agent asks a clarifying question. This happens when a term matches multiple things (e.g., a category and a label both named "Integration"). One short answer and it proceeds.
For "why" questions, start with a search. Try "Show me the top 5 API ideas" first, then "What are customers saying in the comments on the top one?" The agent reads comments to surface the actual pain points, not just the title.
Revenue data requires permission. If revenue figures are missing, your admin account may not have the revenue permission enabled.
Conversations and history
Conversations are saved automatically. Your past chats appear under Recents in the history sidebar, where you can resume any past thread, rename conversations, or delete them — and when you resume a conversation, the agent retains the context from that thread. Click New chat to start fresh, or choose Open in full page to give the agent more room.
Conversations are private to you; there is no conversation sharing between admins. There is no conversation limit during the beta period, though we may bulk-remove conversations at certain points as we make larger changes.
Access, data & privacy
An account owner turns it on for your account in Settings → Integrations → AI Assistants. See Turn the UserVoice agent and MCP on or off. For questions about which AI model powers the agent, data handling, and tenant isolation, see the UserVoice agent FAQ.
For questions about availability and invites, which AI model powers the agent, data handling, and tenant isolation, see the UserVoice agent FAQ. You can also use the agent's tools from Claude, ChatGPT, or other AI tools through UserVoice MCP (Beta) — see Connect UserVoice to Claude, ChatGPT, or Codex.
For common questions — access, model, privacy, and quick "can it do X?" answers — see the UserVoice agent FAQ.