Your agents drive. Your customers steer.

Five lanes. One loop.

TheyAsked runs the customer side. Your agent writes the code. You stay in the one seat that matters: the review.

1. 01 Ask (Customers)

   • Public portal
   • In-app widget
   • iOS and Android
   • Slack and support
   • API and imports

2. 02 Weigh (TheyAsked)

   • Merge duplicates
   • Rank by votes, MRR, momentum
   • Write the brief
   • Check your guardrails

3. 03 Build (Your agent)

   • next_request
   • get_brief
   • Writes the code
   • Opens a pull request

   Claude Code · Cursor · Codex · any MCP client

4. 04 Review (You)

   • Read who asked and why
   • Approve and merge

5. 05 Tell (TheyAsked)

   • Status moves to done
   • Changelog published
   • Email and in-app to all who asked

Watch what happens when they ask.

One request, start to finish. It arrives, it’s weighed by who asked and what they pay, it ships, and everyone who asked hears back. Real interface, demo data.

Hand over the wheel when you’re ready.

1. L1 Suggest

   TheyAsked ranks what to build. You decide and do the rest.

2. L2 Draft

   Briefs, changelog entries and customer emails are written for you to approve.

3. L3 Drive

   Requests that pass your rules go straight to your agent. You only review the code.

Shipping is half the job.

One line.

claude mcp add --transport http theyasked https://theyasked.me/mcp

Works the same in Claude, Cursor, Codex and any MCP client. API and webhooks for everything else.

Everything they asked comes with you.

Votes, voters, comments and dates come along, so nobody has to ask twice. Free on every plan.

• Upvoty
• Canny
• CSV
• XLSX
• TheyAsked backup

Switch in minutes · How importing works

Use cases

• Public feature board: One public place where customers post, vote and see what you decided.
• In-app feedback widget: Requests arrive from inside your product, already signed in, with the account attached.
• Vote or request an integration: Logos, a Vote button and “Request another…” right on your integrations page.
• Feature voting inside a mobile app: A modal sheet in your iOS or Android app, with back handling and no push setup.
• Public roadmap: Planned, In progress, Shipped: customers stop asking “is this happening?”.
• A changelog that tells the people who asked: “You asked. We shipped.” lands with exactly the people who voted.
• Validate demand before you build: Put three candidates in front of real accounts and read the revenue behind each.
• Family and consumer apps: Parents and kids vote; each request counts once per family, revenue once per plan.
• B2B: prioritise by revenue: Rank requests by the MRR behind them, the renewals at risk and the segment asking.
• Ask Claude about your feedback: “What should we build next?” answered in Claude with charts from your own data.
• Migrate from another tool: Requests, votes, voters and comments come over with their dates. Re-run any time.
• Customer success and sales: Log what an account asked on a call, on their behalf, and tell them when it ships.
• Schools and classes: Teachers, students and parents ask; you see which classes and schools want it.

Free for product managers

• MCP server for customer feedback with Claude, step by step
• Prioritize Feature Requests by Revenue With One Formula
• Self-driving software and how AI agents decide what to build
• ChatGPT prompts for product managers (12 free prompts)
• Claude skill for feedback triage (free SKILL.md)
• Claude skill for release notes from requests (SKILL.md)
• Feature Request Prioritization Scorecard (Free CSV)

Give your software somewhere to go.

Start free · Read the docs

Pricing: Free $0, Pro $29, Business $99 a month. See plans.