Funnel Analysis vs Path Analysis: When to Use Each
Funnel analysis vs path analysis comes down to one question: do you already know the steps? Funnel analysis measures how many users move through a fixed sequence of steps you defined in advance, and shows where they drop. Path analysis maps the routes users actually take, including detours, loops and exits you did not plan for. Use funnel analysis to find where onboarding leaks. Use path analysis to find out what users do instead.
For a product manager fixing onboarding drop-off, the practical order is usually funnel first, path second. The funnel tells you which step loses the most users. Path analysis then tells you what those users did before they left, or where they went instead. This article explains both methods, compares them side by side, and covers retention funnels and funnel tracking for apps.
What funnel analysis measures
A funnel is an ordered list of events. For a subscription app it might be: install, onboarding start, quiz complete, paywall view, trial start, first paid renewal. For each step you count the users who reached it, then calculate step-to-step conversion and overall conversion.
You get two things from this. First, a ranked list of the leakiest steps. Second, a way to compare segments: paid versus organic traffic, iOS versus Android, one country versus another. To compute the numbers from your own data, divide users at step N by users at step N-1 for step conversion, and users at the last step by users at the first for overall conversion. Use the same time window and the same user population for every step, or the ratios will not be comparable.
The limits are built in. A funnel only shows what you defined. If users skip a screen, go back, open settings, or leave and return the next day, a strict funnel either hides that behavior or counts it as a drop. It tells you where, not why.
What path analysis measures
Path analysis, sometimes called flow or journey analysis, starts from the data instead of a hypothesis. You pick an anchor event, such as onboarding start or the last screen before churn, and look at what users did before or after it. The output is usually a diagram of the most common sequences, with branches for less common ones.
It answers questions a funnel cannot:
- What do users do right after they see the paywall and do not buy?
- Which screens do users visit between install and the first quiz answer?
- Are there loops, such as users going back and forth between two screens?
- Which route do users who convert take, and is it different from the route you designed?
Path analysis is exploratory. It generates hypotheses, and it gets noisy when the app has many screens or when event naming is inconsistent. It also does not give a clean conversion rate by itself, because there is no single denominator.
Funnel analysis vs path analysis: side-by-side comparison
Starting point
- Funnel analysis: a defined sequence of steps you chose in advance
- Path analysis: one anchor event, with the routes before or after it discovered from the data
Main question
- Funnel analysis: where do users drop, and how much does each step lose?
- Path analysis: what do users actually do, and where do they go instead?
Output
- Funnel analysis: step conversion rates and overall conversion, comparable across segments
- Path analysis: a diagram of common sequences, loops and exits
Best for
- Funnel analysis: prioritizing which step to fix, and measuring whether a change worked
- Path analysis: forming hypotheses about unexpected behavior and finding missing steps
Main weakness
- Funnel analysis: hides everything outside the defined steps
- Path analysis: noisy with many events, and gives no tidy conversion rate
Data requirement
- Funnel analysis: reliable events for each step, with consistent user identity
- Path analysis: a broader and cleaner event log, including navigation and exit events
Which one to use for onboarding drop-off
Start with a funnel. Define the onboarding steps as they exist today, split the funnel by acquisition source and platform, and find the step with the largest absolute loss. Fixing the biggest leak usually beats polishing a step that already converts well.
Then switch to path analysis on that step. Anchor on the screen where users drop and look at the previous few events and the next few events. Typical findings include:
- Users open a help or settings screen before leaving, which suggests the step asks for something they did not expect.
- Users repeat the same screen, which suggests a validation error or confusing input.
- Users who convert skip an optional screen that others spend time on, which suggests the optional screen adds friction.
Each of these becomes a testable change. Run the change as an experiment and go back to the funnel to read the result. If you want a refresher on running tests properly, see our guide to A/B testing. This loop, funnel to locate, path to explain, funnel to verify, is the core workflow.
Retention funnels: using both methods after the first session
A retention funnel applies the same idea to time after signup. Instead of screens, the steps are milestones: first session, key action completed, second-day return, first renewal. The conversion between steps is a retention rate for that milestone, and you can cut it by cohort, which is the group of users who started in the same week or came from the same campaign.
To compute it from your own data, take a cohort, count how many users hit milestone one, then how many of those hit milestone two within your chosen window. Keep the window fixed across cohorts. If you change it halfway, you will see a trend that is only a definition change.
Path analysis helps here too. Anchor on the key action that predicts return in your product, and look at the sequences that lead to it. Compare them with the sequences of users who never reach it. This often shows a first-session behavior worth guiding new users toward. Be careful with causality: a path that correlates with retention may just reflect users who were already motivated.
Funnel tracking in an app: what to instrument
Good analysis depends on events you can trust. For funnel tracking in an app, a few rules prevent most problems:
- Name events consistently and keep a written tracking plan, so that one step does not exist under two names.
- Track step views and step completions separately. A view without a completion is a drop; a completion without a view is a tracking bug.
- Carry one user identifier across the whole journey, including web and app, or the funnel will break at the handoff.
- Send acquisition parameters such as UTM values along with the first event, so you can split funnels by source.
- Record exit and navigation events, such as back taps and app backgrounding, if you plan to use path analysis.
The web-to-app handoff deserves special attention. In a web2app funnel, the user sees ads, a landing page or quiz, a personalized plan and a paywall in the browser, pays on the web, then installs the app and logs in. Analytics that only live inside the app miss the first half of the journey. Where web payments are allowed, the rules differ by region and change over time, so check the current App Store Review Guidelines and Google Play policy, and see our web2app funnel compliance guide for the background.
Tools you can use
Any product analytics tool that supports event funnels and flow or path views will do the job. Amplitude, Mixpanel and PostHog are common choices, and warehouse-based setups work too if your team prefers SQL. The tool matters less than the tracking plan behind it.
web2wave is one practical option for the web part of the journey. It is a no-code builder for web2app funnels, so quiz steps, paywall screens and checkout are built in one place, and each step maps naturally to a funnel stage. It connects to Amplitude, PostHog and Mixpanel, and you can see the options on the integrations page. For how the built-in numbers work, read the analytics post. If your drop-off sits in the quiz, the quiz builder guide covers how step design affects completion, and the paywall builder guide covers the screen where many funnels lose the most users.
If you are comparing builders, our guide to web2app funnel builders lists the main options.
FAQ
What is the difference between funnel analysis and path analysis?
Funnel analysis measures conversion through a fixed, ordered set of steps and shows where users drop. Path analysis discovers the sequences users actually follow around an anchor event, including detours and loops. One tells you where the loss is, the other helps explain what users did instead.
Should I run funnel analysis or path analysis first?
Run the funnel first. It shows which step loses the most users, so you know where to look. Then use path analysis on that step to generate hypotheses about the cause, and return to the funnel to measure the result of any change.
What is a retention funnel?
A retention funnel is a funnel whose steps are milestones over time, such as first session, key action, return visit and renewal. You calculate it by cohort, using a fixed time window, so each step shows the share of a cohort that reached the next milestone.
What do I need for funnel tracking in an app?
You need consistent event names, separate view and completion events for each step, one user identifier that persists across web and app, and acquisition parameters attached to the first event. Without these, funnels show gaps that are tracking errors rather than real drop-off.
Can path analysis replace funnel analysis?
Not for most onboarding work. Path analysis does not give a clean conversion rate for a defined journey, and it is hard to use for judging an experiment. It works best as a companion to funnels, used to explain a drop that the funnel has already located.
If you want to see how web2app funnels and their analytics fit together, you can book a demo or start free.
.webp)
