> For the complete documentation index, see [llms.txt](https://help.usejimo.com/docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://help.usejimo.com/docs/analyze/analytics.md). # Analytics How Jimo counts views, completions and drop-offs, which number answers which question, and what to do when users stop halfway through an experience. Every experience comes with its own analytics: who saw it, who finished it, where people stopped, and whether it changed what they did next. This page explains how each number is counted, so you can read them without guessing and act on what they say. Open any experience from its list: its analytics sit at the bottom of the experience page, in the **Statistics** and **Issues** tabs. Every number can be filtered by date range.

Statistics tab of a tour: A the Statistics and Issues tabs, B the date filter, C Went through all steps, D Users reached, E Total number of clicks, F the Goal card, which invites you to set up a goal, then the step histogram below

## How Jimo counts Most confusion on analytics comes from three words that sound alike: reached, started and completed. They are not the same population.

Three bars of decreasing width: Users reached saw the first step, Started began the experience, Completed reached the last step or clicked a Complete tour action; each count is unique per user

| Term | What it counts | Unique per user | | ------------------ | ----------------------------------------------------------------------------------------------------------------------------------------------------------------- | --------------- | | **Users reached** | Users who saw the first step of a tour, survey or banner, or at least one hint of a hint group | Yes | | **Started** | Users who began the experience. For a survey, users who answered at least one question | Yes | | **Completed** | Users who reached the end: the last step of a tour, or a **Complete tour** action on a button (a Go to step option); a submitted survey; all tasks of a checklist | Yes | | **Step views** | Users who saw a given step | Yes | | **Views** (graphs) | Every display, including repeat displays to the same user | No | #### What counts as a view A view is a display of the experience to a user, not a click. For a tour or a survey, it is the first step appearing on screen. For a hint group, it is any one of its hints. For a checklist, it is the checklist being seen at least once. For a Resource Center, it is the user opening it. #### Why the graph can be higher than the KPI The KPI card counts each user once. The daily graph behind it counts every display. A user who sees a banner on three different days adds one to **Users reached** and three to the graph.

One user sees a banner on Monday, Tuesday and Wednesday: the Users reached card adds 1, the daily graph behind it adds 3

#### Completion rate Completion rate is completed users divided by the users the experience counts from: users reached for a tour, users who started for a survey, users who viewed it for a checklist. A survey completion rate is often higher than you expect, because people who never answered the first question are left out. #### Drop-off Drop-off is the share of users who saw a step but not the next one. Step 1 is measured against the users who started. A step with a 50% drop-off lost half the people who reached it, whatever the size of the tour. #### Clicks The **Total number of clicks** card counts clicks on buttons that do something beyond moving the experience forward: Open Post, Navigate To, Launch Experience, Run JavaScript Code, Open Calendar. Buttons that only go to the next step or close the experience are not counted. For the exact KPIs of each experience type, see [Metrics by experience type](/docs/analyze/analytics/metrics-by-experience-type.md). ## Which number answers which question Start from your question, then read the one number that answers it. | Your question | The number to read | Where | | ------------------------------------------- | ----------------------------------------------------------------- | --------------------------------------------------------------------------------------- | | Did people see it? | Users reached (Unique viewers for hints) | Statistics tab | | Did they finish it? | Went through all steps, Finished submissions, Completed checklist | Statistics tab | | Where do they stop? | Step histogram, biggest drop-off | Statistics tab, tours with 2 steps or more | | Did they click what matters? | Total number of clicks | Statistics tab | | Did it change what they did next? | Reached goal | Statistics tab, once a [Goal](/docs/publish/mesure-outcomes-goal.md) is set | | Which version works better? | Conversion rate and confidence per variant | A/B Test block, see [A/B Test](/docs/analyze/a-b-test.md) | | Is it working for one audience? | Reached and completed per segment | Your assistant, see below | | Is something broken? | Steps whose element was not found, hidden, or matched twice | Issues tab | | What did users answer? | Survey report | See [Survey results and exports](/docs/analyze/analytics/survey-results-and-exports.md) | | Do new users get to value? | Activation tracker conversion | [Success Tracker](/docs/analyze/success-tracker.md) | | How many users can I target by my own data? | Identified users | [Users](/docs/analyze/users-and-segments/audience/users.md) | {% hint style="info" %} **Completed is not the same as successful.** A user can finish a tour without doing the thing the tour was about. To measure the outcome, set a [Goal](/docs/publish/mesure-outcomes-goal.md) and read **Reached goal**. {% endhint %} ## Read a funnel For a tour with two steps or more, the step histogram shows how many users saw each step. The first bar is always 100%: it equals the users who started. Each next bar shows how many are left.

Step histogram of a four-step tour: 100%, 38%, 32% and 29%, the biggest drop-off sits between step 1 and step 2 — Example funnel, illustrative numbers

Read it in three passes. ### The shape A funnel that declines a little at every step is normal: some users always leave. A funnel with one cliff, where most users leave at the same step, points at that step. ### The biggest drop-off The biggest drop-off is the step that loses the most users in number, not the highest rate. A 60% drop on a step seen by 5 people matters less than a 30% drop on a step seen by 400. For example, in the four-step tour of the figure above, seen by 100 users, 38 reached step 2: 62 users, 62%, left at the first transition. Steps 3 and 4 kept most of those who got there, 32 then 29. The problem is step 1 or the move to step 2, not the end of the tour. ### Small numbers Below a few dozen users per step, one user changes the percentages a lot. Wait for more data before you rewrite a step. ## What to do with a drop-off Work through the likely causes in this order, from the cheapest to check to the most expensive to fix.

Five causes of a drop-off to check in order, from the cheapest to check to the most effort to fix: the step could not show, it asks for something not possible yet, wrong moment, too much copy, audience too broad

#### The step could not show Open the **Issues** tab first. If the step's target element was not found, was hidden, or matched several elements, users never saw that step, and the funnel stops there for a technical reason. Fix the element, see [Targeting on dynamic apps](/docs/build/positioning/targeting-on-dynamic-apps.md).

Issues tab of an experience: A the Issues tab, B the Troubleshoot button, C the user, D the step concerned, E the issue, such as Could not find element with selector, F the URL where it happened

The **Issues** tab shows an **Issue reporting** table with the columns **User**, **Step**, **Issue** (for example "Could not find element with selector ..."), **URL** and **Date**, and a **Troubleshoot** button above it. Each row is one occurrence: one user, on one step, on one page. #### The step asks for something the user cannot do yet A step that waits for a click on an element on another page, or for a form to be filled, loses everyone who is not there. Check what moves the user to the next step in the [Steps Triggers](/docs/build/triggers-and-conditions/steps-triggers.md). #### The experience shows at the wrong moment A tour that opens while users are in the middle of a task gets closed. Review its trigger in [Show on](/docs/publish/trigger-on-the-right-spot-show-on.md), its timing in [When](/docs/publish/trigger-at-the-right-time-when.md) and its pages in [Where](/docs/publish/target-specific-urls-domains-where.md). An experience launched by a button action (Launch Experience) checks only its Who: its Where and When do not apply, so look at the step that launches it. If the launched experience uses **No automatic trigger**, only its Goal remains: the experience that launches it decides who sees it. #### The step says too much Long copy and several buttons on the same step slow people down. Shorten it, keep one clear action, and compare the next weeks of data. To compare two versions at the same time, run an [A/B Test](/docs/analyze/a-b-test.md). #### The audience is too broad If users who already know the feature are in the audience, they leave early because they do not need the tour. Narrow it in [Who](/docs/publish/target-right-people-who.md). ## From your assistant With the [Jimo MCP](/docs/integrations/mcp/what-to-ask.md) connected, your AI assistant reads the same analytics and does the comparison for you. > "Where do people drop off in the onboarding tour, and is anything broken on that step?" It returns the step-by-step funnel with the biggest drop-off, then checks the issues reported for the same experience. Other questions that work: "How are we doing this month?", "Compare how Trial Users and Paying Users adopt the new tour", "Is anything broken since our deploy on Monday?". The `weekly_digest` prompt chains them into one weekly summary. See [What to ask?](/docs/integrations/mcp/what-to-ask.md). ## Verify * **Users reached** on the KPI card is lower than or equal to the total of its daily graph. * The **Issues** tab is empty for the experience you are analyzing, or you know why it is not. ## In this section


Metrics by experience type	Every KPI of tours, checklists, hints, banners, surveys and the Resource Center	Metrics by experience type	D2-c02-metrics-by-type.png
Survey results and exports	The Report tab, the users reached list and the CSV exports	Survey results and exports	D2-c03-survey-results.png
A/B Test	Compare versions of a tour on the same Goal	A/B Test	D2-c04-ab-test.png

## Go further


Goal	Measure the outcome, not only the completion	Goal: measure outcomes	B-c05-goal.png
Success Tracker	Activation and feature usage across experiences	Success Tracker	D2-c05-success-tracker.png
Users & Segments	Read results for one population at a time	Users & Segments	D2-c09-users-and-segments.png
Changelog analytics	The analytics of changelog posts, on their own dashboard	Analytics	cover-changelog.png

**Next page:** [A/B Test](/docs/analyze/a-b-test.md)