Use this when: You have users but no clear picture of where they drop off between first touch and value delivery. You know "conversion is low" but cannot say exactly where or by how much.
You're done when: You have a live funnel dashboard with conversion rates at each stage, you have identified the biggest leak, and you have set an improvement target with a timeframe.
The Sequence
Benchmarks
Template
What a funnel actually is
A funnel is a sequence of stage transitions from first touch to value and, eventually, revenue. It answers one practical question: where are people dropping out?
Good funnels have four properties:
- Ordered stages. Step 2 happens after step 1. If users can do the steps in any order, it is not a funnel.
- Clear stage boundaries. Each stage is defined by one specific event or state change, not a vague idea like "engaged."
- Consistent grain. Pick one unit and stick to it: visitor, signup, account, workspace, or deal. Do not mix user-level and account-level stages in the same funnel unless that is explicitly the goal.
- Actionable steps. Every stage should point to an owner and a lever. If a stage drops, someone should know where to look.
If a stage cannot be measured clearly or improved by a team, it does not belong in the funnel.
Define the stages before you build the chart
Most startup funnels use some version of these stages:
- Acquisition. Someone arrives or enters the pipeline.
- Activation. They reach the first meaningful value moment.
- Engagement. They use the product beyond the first touch.
- Retention. They come back after a defined period.
- Revenue. They pay, expand, or renew.
- Referral. They bring in someone else.
You do not need all six. Most funnels should have 4-6 stages max.
Concrete examples:
- Self-serve SaaS: visitor → signup → first project created → active in week 4 → paid
- Sales-led SaaS: lead → qualified → demo completed → pilot/live → closed won
- Marketplace: visitor → account created → first listing/request → first match/booking → repeat transaction
The key is that each stage must represent a real transition:
- Bad: signup → engaged → power user
- Good: signup → created first project → invited teammate → active in week 4
Activation is the most important stage to define well. It is not "logged in" or "visited dashboard." It is the first moment the user gets the value they came for. If you define activation too loosely, the whole funnel becomes misleading.
Fix the Biggest Leak First
The math on this is simple and almost universally ignored. If your funnel looks like this:
- Visitors → Signup: 8%
- Signup → Activated: 15%
- Activated → Retained (month 1): 70%
- Retained → Paid: 5%
Your activation step is the biggest leak. Improving it from 15% to 25% (a 67% relative improvement) means 67% more users flowing into retention and revenue. That single change has more impact on revenue than doubling your retention rate from 70% to 100% (impossible) or doubling your paid conversion from 5% to 10%.
Teams default to working on the stage they understand best. If you are a growth marketer, you optimize acquisition. If you are a product designer, you optimize onboarding. If you are a sales team, you optimize conversion. The funnel does not care about your job title. It cares about the math. Fix the biggest leak, regardless of which team owns it.
The benchmarks above help you calibrate. If your signup-to-activation rate is 15% and the average is 20-30%, you have clear room. If your month-1 retention is 80% and the average is 60-70%, that step is actually performing well. Focus elsewhere.
Example
A developer tools startup had 4,000 monthly signups and $12K MRR. They built their first funnel and found: 4,000 signups, 600 completed setup (15%), 400 created their first integration (10% of signups, 67% of setup), 320 still active at month 1 (80% retention), 48 paid (15% of retained). The biggest leak was setup completion at 15%. They dug into session recordings and found that step 3 of 5 in their setup wizard required an API key that most users did not have ready. They added a "skip for now" option and a guided path to generate the key. Setup completion went from 15% to 38% in three weeks. That single change, with no increase in traffic, added $18K in ARR over the next quarter. They never touched acquisition.
Written with ❤️ by a human (still)