How it Works

Five steps. From funnel chaos to shipped GTM agents.

Most GTM documentation lives in slides and spreadsheets and stays out of date the day after it's written. Quokka makes the funnel a living artifact — and turns it into the agents your team actually ships.

01

Lay out the funnel the way it actually runs.

Drop the steps. Awareness through close. BDR sequences, AE handoffs, pre-sales POCs. Parallel branches that merge later are first-class — the math handles them.

Each step is a node with a name, an owner (BDR, AE, SE, RevOps), a duration, and a description. Double-click to rename. Drag a node next to another to auto-connect. ⌘+N adds a step, ⌘+D duplicates one, Delete removes it.

02

Tag the GTM stack once. It's tagged everywhere.

Salesforce, Outreach, Gong, Clearbit, Apollo — tag every step with what powers it. Quokka keeps a global library across all your flows so the same tool autocompletes everywhere.

The library shows usage counts: which tools touch the most steps, which are only used once, which are stranded after a rep churns. The first honest audit of what your reps actually use.

03

See critical path from MQL to closed-won.

Sequential steps sum. Parallel branches that converge take the max. Quokka highlights the critical path — the actual gating chain that dictates total cycle time.

Total recomputes live as you edit durations. Hover a terminal node to see the time-to-here for any path. Cycles and orphan steps surface as warnings in the top bar. You stop guessing what's gating the deal.

04

Score every step for agent leverage.

Flag the steps where an AI agent has real leverage — repetitive, well-instrumented, high-volume. Lead enrichment scores high. Reading the room in a discovery call doesn't.

Per-step readiness scores with reasoning. Top-3 candidates by impact. An orchestration blueprint showing which agents to build, what they trigger on, and where humans still gate.

05

Ship the agent (or preview it first).

Today: 'Try this with sample data' — Quokka role-plays each recommended agent against synthetic input so you see exactly what it would produce. Next: deployed agents that fire on real triggers.

The downloadable AI readiness report includes every agent's trigger, inputs, outputs, and human checkpoints — paste it into Notion, hand it to engineering, or use it as the spec for whoever's building the bot.

Under the hood

The mechanics, briefly.

The math

Critical path on a DAG

Topological sort over your steps, longest-path DP to compute total time, backtrack to highlight which steps gate everything. Cycle detection in the top bar so you know if your graph is broken.

The data

Your flow, your data

Every flow saves to your account in our Postgres database. Auto-save on every edit. Sign-in is Google OAuth — no passwords for us to lose. JSON export gives you a portable copy any time.

The stack

Built to feel like a tool

React Flow on the canvas, server-rendered marketing for fast loads, dark mode by default. Designed for executives and operators, not for tutorials.

Build your first map.

Sign in with Google. Start from a blank canvas. Hand it to your team when you're done.

Start mapping