The Userflow MCP Prompt Playbook: Workflows by Team

A PM reading a launch and a CS lead prepping a renewal look at the same product data and ask completely different questions. Same numbers, different job, different phrasing entirely.

Userflow MCP is a Model Context Protocol server that connects your Userflow product data to AI tools like Claude, ChatGPT, and Claude Code, so anyone can ask about flows, NPS, segments, and user behavior in plain language. Our 50+ prompt library sorts those questions by task. This playbook sorts them by team.

Copy your section, send your CS lead theirs, and MCP stops being a tool one person found and becomes how the team answers questions.

Key Takeaways

• Userflow MCP benefits every team, not just the one that owns analytics. PMs, CS, marketers, and ops all ask product questions. Each gets a section here.
• These prompts are workflow-shaped. They map to how a role actually spends its week, and many are multi-step so you can follow up in the same conversation.
• Every section has a "combine with" move. Once you've connected more than one MCP, a single prompt can pull Userflow data alongside Slack, or Mixpanel.
• Built to forward. Find your team, copy the section, drop it to your lead. That's the whole point.
• For the full reference, use the library. This is the shareable companion, not a replacement.

How to Use This Playbook

Find your team below. Each section has five or six prompts, one pro tip, and a "combine with" suggestion for when you've connected more than one tool.

Two things make every prompt land better. Be specific: name the flow, the segment, and the time range instead of asking in the abstract. And iterate in the same conversation: start broad, then follow up with "show me the quotes" or "break that down by segment" to drill in. The prompts below are starting points. Your flow names and segments make them yours.

New to setup? Start with the Userflow MCP setup guide.

For Product Managers

Product managers use Userflow MCP to check flow health, read launches, and walk into sprint planning with the drop-off numbers already in hand.

The week runs on the same handful of questions. Here they are as prompts.

• "Which onboarding step lost the most users this week, and how does that compare to last week?"
• "Rank our live flows by completion rate and flag any that dropped more than 10 points since the last release."
• "For the segment that saw [New Feature Tour], what share activated within seven days?"
• "Summarize how the new activation flow is performing against the version it replaced."
• "Before I set next sprint's goals, show me the three flows with the worst step-level drop-off."
• "Which flows are live in production but haven't been edited in 90 days?"

Pro tip: Turn the Monday review into a saved Claude Project prompt. The standup number is ready before the meeting starts, not built during it.

Combine with Mixpanel MCP: Tie a flow to the downstream behavior it's supposed to drive, in one prompt. "From Userflow, pull the users who completed [Activation Tour]; from Mixpanel, show how many came back in the next seven days." Product Adoption Insights tells you they finished. The cross-tool prompt tells you whether the ones who finished actually stuck around.

For Customer Success

Customer success uses Userflow MCP to prep for calls without logging into anything. Account activity, NPS, and open adoption gaps, pulled in the two minutes before the meeting.

• "Give me a two-minute account summary for [Account]: recent flow activity, NPS, and any flagged events."
• "Which of my accounts stalled in setup and haven't come back in the last 14 days?"
• "Did anyone at [Account] see the [Feature Announcement] we shipped last month?"
• "Summarize NPS detractor themes across my book of business this quarter."
• "List accounts that finished onboarding but never triggered [Key Value Event]."
• "Which accounts in the [Churn Risk] segment have adoption gaps I should raise on the next call?"

Pro tip: Run the account summary two minutes before every renewal. You walk in knowing exactly what that account has and hasn't touched, instead of guessing.

Combine with Slack MCP: Pull the list and route it in one motion. Ask for the accounts that stalled in setup this week, then have your AI tool post those names to your CS channel in Slack. The prep and the handoff land in one thread instead of three tabs.

For Marketing

Marketing uses Userflow MCP to see who a campaign actually reached and to pull real customer language before developing the next one. Who a campaign reached, what customers actually said, and which ones to ask for a review, pulled straight from product data

• "How many users in [segment] saw the [Feature Announcement] we shipped last week, and how many acted on it?"
• "Pull NPS promoters from this quarter so I can shortlist customers for case studies and reviews."
• "Summarize the open-text responses to [Survey] so I can lift real customer language for the campaign."
• "Which users completed [Onboarding Flow] but never reached [Key Feature], so I can target them with an education campaign?"
• "How did the in-app [Announcement] perform against the last one we ran?"
• "Which segments grew fastest this quarter, so I know where to point the next campaign?"

Pro tip: Before you build the next campaign, pull the exact words customers used in [Survey] responses. Real phrasing from real users lands harder than anything drafted in a doc.

Combine with Mixpanel MCP: Tie an in-app announcement to the conversion it was supposed to drive. "From Userflow, show who saw the [Feature Announcement]; from Mixpanel, show how many of them completed [Target Action] in the next two weeks." Reach is the first number. Whether it moved anyone is the one that tells you if the campaign worked.

For Ops and Leadership

RevOps, ops, and execs use Userflow MCP for the wide-angle view: segment growth, portfolio health, and the adoption numbers a board deck needs.

• "How has the [New Signups] segment grown over the last quarter?"
• "What share of our user base sits in the [Churn Risk] segment right now?"
• "Summarize the NPS trend over the last two quarters for the board deck."
• "List the top 10 accounts by flow-completion volume this quarter."
• "How many flows are live in production, and how many are stale drafts?"
• "Which flows are still live but haven't been edited since January?"

Pro tip: Build one quarterly "state of adoption" prompt for board prep. The question that used to need its own dashboard becomes a paragraph you can paste into the deck.

Combine with Mixpanel MCP: Put onboarding health next to long-term stickiness for the board view. "From Userflow, show activation and NPS for [segment]; from Mixpanel, show 30-day retention for the same users," so what happens in onboarding and what happens after it sit in one answer.

One Playbook, One Place Your Team Already Works

The point of sorting by team isn't neatness. It's adoption. A prompt library only changes how a company works if the prompts reach the people who'd use them, and people adopt what fits their week, not what sits in a doc they never open.

So find your section, and send someone else theirs.

Frequently Asked Questions

How is this different from the 50+ prompt library?
The 50+ prompt library organizes prompts by job-to-be-done: daily performance, onboarding diagnostics, NPS analysis, and so on. This playbook organizes them by team, so you can hand a colleague the exact section that fits their role. Use the library as your full reference; use this to get MCP into your team's hands.

Which teams get the most out of Userflow MCP?
Any team that asks product questions. Product managers use it for flow health and launch reads, customer success for renewal prep and account lookups, and ops or leadership for segment growth and board numbers.

Which AI tools do these prompts work in?
Userflow MCP works with Claude, ChatGPT, and Claude Code. The prompts in this playbook work the same way across them.

Can I customize these prompts for my team?
Yes, and you should. Swap in your real flow names, segment names, and time ranges. A prompt that names "the Welcome Tour, last 30 days" returns a sharper answer than one that asks in the abstract.

How do the "combine with" prompts work?
Once you've connected more than one MCP, a single prompt can reach across them. Reference each source by name so your AI tool doesn't confuse them, for example: "From Userflow, show flow completion for [segment]; from Mixpanel, show retention for the same users." Cross-tool prompts work with any MCPs you've connected, including Slack and Mixpanel.

Ready to Arm Your Team With the Numbers They Need?

Connect Userflow MCP to the AI tools your team already uses, and turn the dashboard hunt into a prompt anyone can run. Start your free trial →