Activation leaks
Find the exact onboarding step where users stop and review sessions from that transition.
Conversion funnel analytics
Conversion funnel analytics that explains every drop-off
Build funnels for signup, onboarding, activation, purchase, forms, and feature adoption. Then use replay, heatmaps, and feedback to understand why users stop.
Problems it solves
Built for teams that need evidence, not another disconnected report
Noisy event definitions
Validate funnel steps with recordings so the funnel represents user progress, not just component renders.
Teams debating causes
Use replay, heatmaps, and feedback to turn a drop-off number into a shared product hypothesis.
Activation funnels
What an activation funnel actually measures
Signup to activated
Measure the share of new signups that reach your activation event, not just the ones that finish onboarding.
Onboarding step drop-off
See which onboarding step loses the most users, then open the recordings of the sessions that stalled there.
Workflow
How PulsePanda turns analytics into product decisions
Name the steps
Create funnels around user intent within a visit: landed, viewed pricing, started signup, signed up, created a workspace.
Measure the leak
Track conversion rate and drop-off across each step, on desktop and mobile.
Explain the behavior
Open recordings, heatmaps, errors, and feedback from the weak transition.
How it counts
How a funnel is counted
Ordered steps, matched one visit at a time, with the replays of the visits that stopped one click from the number.
A funnel in PulsePanda is an ordered list of steps, evaluated one visit at a time. A visit carries across page loads and across new tabs opened within 30 minutes, so a signup that starts on the pricing page and finishes in a second tab is still one journey.
Within each visit, events are put in time order and the steps are matched in sequence. A visit has to reach step one before step two can count, but other things can happen in between: a visitor who reads the docs between starting signup and finishing it still converts. Each visit counts once per step, however many times it repeats one.
Every step reports three numbers:
- Of entrants: the share of visits that reached step one and got this far. On the last step, that is the funnel's conversion rate.
- From previous: the share of visits at the previous step that went on to this one. This is where a single weak transition shows up.
- Dropped: how many visits reached the previous step and stopped there.
The step that loses the most visits is called out as the largest drop, and a sample of the visits that stopped is listed with a link straight to each session replay, so “why here?” is one click from the number.
You can look at the last 7, 30 or 90 days and split any funnel into desktop and mobile visits. That split is usually the first one worth making: a form that works on a laptop and fails on a phone looks like one leaky step until you separate them.
Steps
Four kinds of funnel step
Mix them in one funnel. The most useful funnels combine pages reached with one or two events.
| Page URL | The visit reached a page whose URL contains, or exactly matches, the text you give: /pricing, /signup, /onboarding. It needs no instrumentation, and the page a form leads to is usually the most honest sign that the form worked. |
|---|---|
| Event type | Any event of a type the script already records: a click, an input, a rage click, a JavaScript error, a feedback submission, an announcement view. Useful for steps like “hit an error” or “gave feedback”. |
| Defined event | A click event you defined from autocaptured clicks, by element id, class, text or tag. The definition applies to clicks recorded before you made it, so a new funnel can use last month's data. |
| Tracked event | A named event you send with PulsePanda.track(), such as workspace_created or plan_upgraded. The right step for outcomes that have no page of their own. |
What you get
Connected analytics capabilities in one workspace
| Signup funnels | Understand where trial users abandon account creation or workspace setup. |
|---|---|
| Activation funnels | Track whether users reach the first valuable product moment. |
| Form funnels | Find validation, trust, and layout problems that block submissions. |
Examples
Three funnels to build first: signup, onboarding and activation
Most SaaS teams need the same three funnels before any others. Each is written as the steps you would enter.
| Funnel | Steps | What a leak usually means |
|---|---|---|
| Signup (conversion) funnel | Page /pricing → page /signup → page /welcome | Pricing to signup: the plan or the price is unclear. Signup to welcome: the form is failing, often on mobile or on one validation rule. |
| Onboarding funnel | Page /welcome → tracked workspace_created → tracked teammate_invited | A step nobody finishes usually asks for something the user does not have yet: a teammate's email, a data source, an API key. |
| Activation funnel | Page /dashboard → tracked project_created → tracked report_viewed | Users who set up and never reach the first useful result. This is the funnel that shows churn before it happens. |
An onboarding funnel measures whether people finish the setup you designed. An activation funnel measures whether they reach the moment the product is useful. The two overlap, and the gap between them is where most early churn hides. If you have not decided what “activated” means yet, activation funnel benchmarks covers how, and what an onboarding funnel is walks through the steps.
Steps are matched within one visit, so these funnels describe what happens in a sitting. When activation normally takes several days, read the funnel alongside cohort retention.
Pitfalls
Reading funnel drop-off without fooling yourself
- A first step that matches everything. “URL contains
/” is true of every page, so the funnel starts with every visit on the site. Start from the page where the journey really begins, or use an exact match. - Clicks standing in for outcomes. A click on “Create workspace” is not a workspace. If the request fails, a click step still counts it. Use the page that success leads to, or a tracked event sent once it has worked.
- Steps in the wrong order. Steps are matched in sequence, so a step that can happen before the previous one, like reading pricing after signing up, undercounts. Order steps the way the journey must happen, not the way you hope it does.
- Desktop and mobile blended together. A funnel that converts at 30% may be 45% on desktop and 10% on mobile. Split it before deciding which step to fix.
- Small numbers. Forty visits reaching a step is too few to compare one week with the next. Use a longer period until each step has a few hundred.
- Rates without sessions. A drop-off rate says where. Only the sessions say why. Watch five of the visits that stopped before changing the step.
FAQ
Questions teams ask before switching
What is conversion funnel analytics?
Conversion funnel analytics measures how users move through a sequence of steps, such as signup, onboarding, activation, purchase, or form submission.
Why pair funnels with recordings?
Funnels show where users drop off. Recordings, heatmaps, errors, and feedback help explain why the drop-off happened.
Can PulsePanda track SaaS activation funnels?
Yes. PulsePanda is built for SaaS teams that need to track activation steps and connect those steps to user behavior evidence.
FAQ
Activation and onboarding funnel questions
What is an activation funnel?
An activation funnel counts how many new users reach the first moment your product is useful to them, such as creating a first project and viewing a first report, and shows the step where the rest stop. It differs from an onboarding funnel, which measures whether people finish the setup steps you designed.
What is an onboarding funnel?
The sequence of setup steps a new user goes through after signing up, such as creating a workspace, inviting a teammate and connecting data. Its drop-off shows which setup step asks too much.
Does a funnel count steps across visits?
No. Steps are matched within one visit, which carries across page loads and new tabs opened within 30 minutes. For journeys that take days, read the funnel alongside cohort retention.
Can I use my own events as funnel steps?
Yes. A step can be a page URL, any event type the script records, a click event you defined from autocaptured clicks, or a named event sent with PulsePanda.track().
Can I watch the sessions that dropped off?
Yes. Each funnel lists a sample of the visits that stopped, each linked to its session replay.
Related pages
Explore the rest of the PulsePanda analytics stack
Start understanding your users today
Install once and connect product analytics, replay, heatmaps, funnels, and feedback in minutes.
Start for free