Funnel Analysis for Apps: Find Where Users Drop Out 

What is funnel analysis?

Funnel analysis is an app analytics method for understanding how users progress through a defined sequence of actions toward a goal. Each action becomes a funnel step.

For example, an onboarding funnel might contain the steps App Installed → Seen Onboarding Screen No. 4 → Account Created → First Project Created → Subscription Started.

Funnels are the charts that visualize the conversion rates between these steps. They show how many users reach each stage, how many continue, and where users drop off. This makes funnel analysis especially useful for optimizing complex flows such as onboarding, sign-ups, checkouts, paywalls, and other multi-step processes in your app.

The key metric for funnel analytics is the step conversion rate:

Example: If 800 users download an app and 600 of them complete onboarding, the step conversion rate is 600 ÷ 800 × 100 = 75%. The remaining 25% is the drop-off between those steps.

You can also measure the overall funnel conversion rate by comparing the first and final steps. Funnels, therefore, turn individual analytics events into a measurable user journey. Instead of only knowing how many people used your app, you can see where they succeed, where they get stuck, and which part of the experience you should improve.

Conversion funnel analysis, step by step

A conversion funnel becomes most useful when you map it to a concrete goal in your app. Imagine you have an app with a paid subscription and want to understand how new users become subscribers.

Your funnel contains four steps:

1. Install: 10,000 users. These users install your app. This is the widest point of the funnel and your starting population.
2. Signup: 7,200 users. Of the 10,000 users who have installed, 7,200 create an account. Your step conversion rate is 72%, meaning 2,800 users drop off before completing signup.
3. Core value: 4,500 users. Next, you measure activation. In this example, your app delivers its core value when a user creates their first task. 4,500 of the 7,200 registered users reach this milestone, giving you a 62.5% step conversion rate.
4. Paid: 900 users. Finally, 900 activated users subscribe. That is a 20% conversion rate from first insight to paid and a 9% overall conversion rate from install to paid.

The important part is not just the final 9%. The funnel tells you where you are losing people. Improving signup will have a different effect from optimizing your paywall.

TelemetryDeck funnels help you visualize exactly this progression and identify the steps where users leave a process. For a practical introduction, see our guide to funnel insights or watch our webinar about funnels.

Where funnels find problems: onboarding, checkout, and paywall

Onboarding: An onboarding flow is supposed to move new users toward their first successful experience with your app. A funnel can represent every important step within your analytics data, from opening the app and creating an account to configuring the product and reaching the core value/activation moment. A sudden drop between two funnel steps shows you where new users are struggling (or losing interest).

Checkout: For an e-commerce app, you could create a funnel containing Product Viewed → Added to Cart → Checkout Started → Purchase Completed. If 1,000 users add a product to their cart but only 400 begin checkout, you know where to investigate. Funnels turn a vague problem such as "not enough purchases" into a specific part of the user experience you can improve.

Paywall: Funnels are equally useful for subscription apps. Compare how many users encounter your paywall, start the subscription process, and successfully become paying customers. You can also investigate what users do before reaching the paywall. Combining conversion analysis with retention analytics helps you understand not only who converts, but whether those users keep coming back.

Building an onboarding funnel in TelemetryDeck

Setting up funnels in TelemetryDeck is straightforward. Start by identifying the events that represent the steps in your process. For an onboarding funnel, these might be Account Created, Profile Completed, First Task Created, and Onboarding Completed.

In your TelemetryDeck dashboard, create an Insight and select the funnel chart and funnel query. Then use the visual filter editor to define the condition for each step. Previous conditions automatically apply to the following steps, so you do not need to repeat them.

The finished funnel shows how many users completed each stage and where they dropped off. You can adjust the steps and filters as you learn more about the user journey, making it easy to test assumptions.

For advanced use cases, TelemetryDeck's TQL Funnel Query documentation explains how to extract funnels directly from your analytics data.

Funnel analysis without tracking individual users

Fact: Understanding a user journey does not require knowing who the user is.

TelemetryDeck is built around this principle. Our privacy-first analytics lets you measure how users float through funnel steps without creating identifiable user profiles or collecting personally identifiable information (PII). You can see that an anonymized user completed signup, reached an activation event, and converted without needing their name, email address, or advertising identifier.

This means you can answer important product questions such as "Where do users abandon onboarding?" or "How many activated users become customers?" while respecting user privacy.

The result is the basis for data-driven decisions developers need without turning funnel analysis into surveillance. TelemetryDeck delivers useful behavioral analytics built with the privacy by design principle in mind. Our goal is to help you improve your app based on real usage data without tracking identifiable people (since everyone sleeps better when you don’t have to deal with sensitive data about people’s lives).

FAQ

How many steps should a funnel have?

There is no fixed number, but a good funnel contains enough steps to represent the meaningful decisions in a user journey without measuring every tap (yes, there can be user actions between steps that don’t need to be part of the funnel). For many app funnels, three to six steps provide a useful starting point. If a funnel becomes too detailed, consider splitting it into separate funnels for different parts of the journey.

What's a good funnel conversion rate for an app?

There is no universal good funnel conversion rate. A 10% conversion rate could be excellent for one funnel and a serious problem for another. It depends on the app, business model, acquisition channel, and what the first and final steps represent. Start by establishing your own baseline (that includes adding analytics early). Then compare conversion rates over time (especially every time you launch a new version of your app) and investigate individual step conversion rates to find opportunities for improvement.

Can you do funnel analysis without personal data?

Yes. Funnel analysis requires a way to determine whether the same user progressed through a sequence of events, but that does not mean you need to know the person's identity. TelemetryDeck uses anonymized identifiers to connect events for analytics purposes without collecting personal data. This lets you analyze funnels, conversion, and drop-off while respecting your user’s privacy.