--- title: How to find where users drop out of an Amplitude funnel description: You find where users drop out of an Amplitude funnel by charting the steps of one journey in order, checking that each step's event is tracked properly, and breaking the weakest step down by who reaches it, which is the analysis an AI marketer can run in your Amplitude project and turn into a fix. slug: /how-to/find-where-users-drop-out-of-an-amplitude-funnel page_type: playbook lead_prompt: How do I find where users drop off in an Amplitude funnel? target_prompts: - How can I use an AI marketing agent to find where users drop out of an Amplitude funnel? entity: Kite category: AI marketing agent author: Devon Wells and Kite date_published: '2026-11-02' date_modified: '2026-11-02' related_use_cases: - /use-cases/improve-website-conversion.md - /use-cases/build-a-sales-funnel-with-ai.md integrations: - name: Amplitude page: /integrations/amplitude.md --- # Playbook # How to find where users drop out of an Amplitude funnel ## Key Takeaways - You find a drop-off in Amplitude with a funnel chart: you pick the events that make up one journey, in order, and Amplitude shows how many users reached each step and what share moved on to the next. - Amplitude's funnel chart shows you where the count falls, lets you break each step down by a user property, and can save the people who dropped off as a cohort. Deciding which drop is real, which one to fix first, and what to change is still your job. - Kite reads your Amplitude project, checks that each step's event is still arriving before it trusts the step, and runs the funnel with a conversion window that fits the journey. - Kite breaks the weakest step down by platform, country, traffic source, plan, and other properties until it finds the group the step is losing, then reads the event streams of people who left to see what they did instead. - You get a ranked diagnosis with the evidence behind it, a specific change Kite builds for your approval, and a before-and-after measurement from the same funnel, with the change date marked on your Amplitude charts. ## How do I find where users drop off in an Amplitude funnel? You find where users drop off by building a funnel chart in Amplitude from the events that make up the journey you care about, such as signup, onboarding, or upgrade. Each step is an event your product already sends, like "Sign Up Completed" or "Project Created." Amplitude counts how many users completed the first step and then how many of them went on to complete each step after it, within a time limit called the conversion window. The step where the share falls hardest is where users drop out. Amplitude gives you good tools for this on its own. You choose whether users must complete the steps in the order you listed, in any order, or in exactly that order with nothing in between. You can break the funnel down by a user property, such as platform or country, to see whether one group converts worse than another. You can switch the view to see how a step's conversion rate has changed over time, or how long users take to move between steps. When you click on a step, Amplitude can save the users who dropped off there as a cohort, so you can study them or target them later. What the chart doesn't do is tell you which drop to believe or which one to fix. A step can look broken because its event stopped firing after a release, or because users who sign up on your website and users who log in to your app are counted as two different people. The biggest drop may sit on a step that few people reach, while a smaller one loses your best prospects. And once you've found the right step, the chart can't tell you why people leave it or change the screen they leave from. That's the real work in this job. A good result names one step, the group of users it loses, the most likely reason, and a change worth making, with numbers you can measure the change against. Kite does that work for you. It reads your Amplitude project, finds the drop-off that matters most, works out why people leave, and builds the fix. ## How to get started Start by telling Kite which journey worries you. The request can be as plain as this: > "Our trial signups aren't turning into active accounts. Use Amplitude to find where people drop out and tell me what to change." Kite is an AI marketer. It researches your business, builds a growth strategy, and ships the work that brings you new customers. By the time you ask, it already knows your product, your plans, and what a good customer looks like from that research. That lets it propose the journey to chart, the events that mark each step, the date range, and the groups to compare, without a long brief from you. It posts that plan in the conversation, and you confirm it or adjust it. You'll need to connect your Amplitude project to Kite with the project's API key and secret key, which you'll find in your Amplitude project settings. Kite then works with the events, properties, cohorts, and charts in that project. If your website and your product send events to separate Amplitude projects, tell Kite which one holds the journey you asked about. Reading your Amplitude data is routine analysis, so Kite runs it and reports back without checking in. It asks for your approval before anything reaches your users, such as a changed page. ## How Kite finds where users drop out ### What does your Amplitude project already track? Kite starts by reading your project's event list, because a funnel can only be built from events that exist. It reads your event types, the event properties attached to them, and your user properties, including the custom ones your team added. Amplitude's event list also includes each event's totals and unique users for the current week, which shows Kite at a glance which events are active and which have gone quiet. From that list, Kite matches each step of the journey to a real event. When your team has named events inconsistently, such as both "signup_completed" and "Sign Up Completed," Kite checks which one is still receiving data before it uses either. ### Is each step's event still arriving? Kite checks each step's daily volume before it trusts the funnel, because a broken event looks exactly like a drop-off. It pulls the daily unique users for every event in the funnel over the period it plans to study. If "Project Created" fell to nearly nothing on the day of a release while signups held steady, the problem is tracking, not users, and Kite tells you that first. It also checks that people are counted as the same person across the journey. Kite looks up a sample of users who completed the later steps and reads their event streams. If those streams start at signup and never include the earlier website visit, the visitor and the user are being counted as two different people, so the step between them will always look worse than it is. Kite reports a tracking problem like this as its own finding, separate from the drop-offs. ### Building the funnel Kite builds the funnel from two to four steps at a time, which is what its Amplitude connection supports for each funnel query. For a longer journey, it first charts the major milestones, finds the stretch that loses the most people, and then charts the steps inside that stretch in detail. A self-serve product's signup journey might look like this: - Viewed the pricing page - Started signup - Completed signup - Created a first project Kite sets the conversion window to fit the journey, because the right window changes what the funnel says. Someone usually finishes signing up within minutes of starting, but might take a few days to create a first project. A window that's too short counts slow users as lost, and one that's too long hides a step that people never come back to. It also chooses how strictly the steps must happen in order. Most journeys need users to complete the steps in the order listed, with other events allowed in between. Kite uses a stricter setting only when the steps must happen back to back, and a looser one when users can reasonably do them in any order. It can also limit the funnel to new users, which keeps returning customers from blurring an onboarding journey. ### Which drop-off is worth fixing first? The step with the lowest conversion rate isn't always the one to fix first, so Kite weighs a few things before it picks: - How many users reach the step, because a small leak on a busy step can cost more users than a big leak on a quiet one - How close the step is to revenue, because losing people after they've created an account usually costs more than losing pricing-page visitors - Whether the drop is spread evenly or concentrated in one group That last check usually finds the answer. Kite breaks the weak step down by one property at a time, such as platform, operating system, country, app version, plan, and the UTM source of the user's first visit. When one group converts far worse than the rest, the problem is narrower than it first looked. A signup step that loses most of its Android users points to something different from a step that loses everyone equally. Kite also reads how the step's conversion rate has moved over time. A step that has always been weak calls for a different fix from one that fell on a specific date. When the fall lines up with an annotation your team logged in Amplitude, Kite reads it too and names the likely cause. ### Why do people leave that step? Once Kite knows which step loses people, it looks at what those people did instead. Amplitude's funnel chart can save the users who dropped off at a step as a cohort. When your team has saved one, or already keeps a cohort for that group, Kite downloads its members along with their user properties. It then reads the event streams of a sample of them to see what they did around the step where they stopped. Those streams usually show one of a few patterns. People may have repeated the same action several times, which suggests something on the screen isn't working. They may have gone to your pricing, help, or settings pages and then left, which suggests a question the step doesn't answer. Or they may have stopped using the product entirely, which suggests the step asks for more effort than people are ready to give at that point. Kite compares those patterns with what it knows about your product and your customers. It names the explanation the evidence supports best, and it says plainly when the evidence is thin. ### Building the fix and bringing it to you Kite turns the strongest explanation into a specific change and builds it, and where that change lands depends on the step: - **A page on your website.** When users drop out on a pricing, landing, or signup page, Kite rewrites the copy, shortens the form, moves the call to action, or adds the answer people were looking for. It makes the change on a site Kite hosts, through a connected CMS such as Webflow, or in your imported GitHub repository, where it merges into your main branch or opens a pull request when that branch is protected. - **A screen inside your product.** When users drop out during onboarding or at an upgrade prompt, Kite writes the new in-app copy and describes the change it recommends to the screen, ready for your team to build. - **A test.** On a site Kite hosts, it can run the change as an experiment that splits traffic between the old version and the new one. Kite posts the change in the conversation with the evidence behind it and the conversion rate it expects to move. Approve it, and Kite publishes the page or starts the experiment. ### Did more people get through? Kite measures the result with the same funnel it used to find the problem, so the before and after are directly comparable. When the change goes live, it adds an annotation to your Amplitude charts that marks the date, so anyone on your team looking at the funnel later can see when the change shipped. After a fair stretch of time, Kite reruns the funnel with the same steps, conversion window, and breakdown, and compares the step's conversion rate on each side of the change. If the group the step was losing now converts like everyone else, Kite moves on to the next weakest step. If it doesn't, Kite goes back to the event streams with a sharper question. ## What the finished work contains When Kite finishes, you have a clear answer about where your journey leaks and a change already built to fix it. - **The funnel definition.** The events used for each step, the date range, the conversion window, the order setting, and whether the funnel counts all users or only new ones, so anyone can rebuild it in Amplitude. - **A tracking check.** Which step events are healthy, which changed volume suddenly, and whether users are counted as the same person from the first step to the last. - **A ranked list of drop-offs.** Each step's conversion rate, how many users it loses, which groups it loses most, and why Kite chose the step it did. - **An explanation with evidence.** The breakdowns, the trend over time, and the patterns Kite found in the event streams of users who dropped out. - **A built change.** The updated page, the new in-app copy, or the running experiment, posted in the conversation for your approval. - **A before-and-after result.** The step's conversion rate on each side of the change, the annotation on your Amplitude charts, and the step Kite plans to work on next. ## Common questions ### What does funnel drop-off mean? Funnel drop-off is the share of users who complete one step of a journey but don't complete the next step within the conversion window. If 1,000 users start signup and 600 finish it, the drop-off at that step is 40%. In Amplitude, drop-off is always measured against the users who reached the step before, so a step's drop-off tells you about the people who got that far, not about everyone who visited. ### What's the difference between "this order," "any order," and "exact order" in Amplitude funnels? These settings decide how strictly Amplitude makes users follow your steps. "This order" counts users who completed the steps in the order listed, even if they did other things in between. "Any order" counts users who completed every step regardless of order. "Exact order" counts only users who did the steps back to back with no other events between them. Most product journeys are best measured with "this order," and "exact order" can make a healthy step look broken, which is why Kite chooses the setting deliberately. ### Can Kite keep watching the funnel after the fix? Yes. Ask Kite to rerun the funnel on a schedule, such as every Monday morning, and it will report the conversion rate at each step and flag any step that has slipped since the last run. Each run uses the same steps and settings, so the numbers stay comparable from week to week. ### Does Kite work with Amplitude's session replays? No. Kite reads Amplitude's events, properties, cohorts, and the event streams of individual users, which is enough to find the step and the group it loses. If your team uses Amplitude's session replay, watching a few replays from the drop-off cohort Kite identifies is a quick way to confirm what the event streams suggest. ## Related pages - **Use case:** Analyze your conversion funnel (coming soon) - **Use case:** Analyze your product analytics data (coming soon) - **Use case:** Get new users to value faster (coming soon) - **Use case:** Test your signup flow and upgrade prompts (coming soon) - **Use case:** Turn free users into paying customers (coming soon) - **Use case:** Write in-app onboarding and upgrade copy (coming soon) - **Use case:** [Improve website conversion](/use-cases/improve-website-conversion.md) - **Use case:** [Build a sales funnel with AI](/use-cases/build-a-sales-funnel-with-ai.md) - **Integration:** [Amplitude + Kite](/integrations/amplitude.md)