--- title: Metric Explorer use cases description: "Worked Metric Explorer examples for event analysis, session analysis, and analyses that combine events and sessions, with the setup and how to read each result." product: analytics lang: en last_updated: 2026-10-27 token_estimate: 2755 --- # Metric Explorer use cases > For AI agents: a documentation index is available at [/docs/llms.txt](/docs/llms.txt). Append `.md` to any page URL for markdown, or send `Accept: text/markdown`. Each example on this page answers one product or marketing question with a single [Metric Explorer](https://amplitude.com/docs/analytics/charts/metric-explorer/metric-explorer-build) chart. Every example lists the question, the setup, and how to read the result. Replace the example event and property names with the ones your project tracks. The examples follow three layers of analysis: - **Event analysis**: Several events on one chart, each with its own measurement, filters, group-bys, and smoothing. - **Session analysis**: Visit-based metrics, with sessions defined by what happened in them. - **Events and sessions together**: Event metrics next to session metrics, and hybrid formulas that relate the two. Session examples need session tracking in your project. For setup, go to [Track sessions](https://amplitude.com/docs/data/sources/instrument-track-sessions). In these examples, "all sessions" means `[Amplitude] Any Active Event` measured as **Total Sessions**. For Metric Explorer setups of standard marketing metrics, such as page views per session and exit rate, go to [Marketing metrics recipes](https://amplitude.com/docs/analytics/charts/user-sessions/marketing-metrics-recipes). ## Event analysis ### Sign-ups, purchases, and revenue on one chart **Question**: Is growth coming from more customers, more orders, or larger orders? Set up the chart: - Metric A: `Sign Up`, measured as **Unique Users**. - Metric B: `Purchase`, measured as **Event Totals**. - Metric C: `Purchase`, measured as **Revenue** on the revenue property. - Interval: daily over the last 90 days. Visualization: **Line chart**. - Revenue has a much larger scale than the counts. In **Chart Settings** > **Y-Axis settings**, click **Add Right Y-Axis**, then set metric C to **Right** under **Axis Assignment**. **Read the result**: Compare the slope of each line. When B grows faster than A, existing customers place more orders. When C grows faster than B, the average order is getting larger. Because each metric has its own measurement, this comparison takes one chart instead of a formula or three separate charts. ### Purchases by product category next to total sign-ups **Question**: Which product categories drive purchases, compared with overall sign-up volume? Set up the chart: - Metric A: `Sign Up`, measured as **Unique Users**, with no group-by. - Metric B: `Purchase`, measured as **Event Totals**, with **+ Group-by** on the event property `product_category`. - Visualization: **Line chart**, or **Bar chart** to compare totals for the date range. **Read the result**: Sign-ups stay a single line, and purchases split into one line per category. The breakdown table lists the categories, and the Top N control above it sets how many the chart draws. If the category sits inside a cart (array) property, group by the cart property and choose **Counting Event** or **Counting Item**. For the difference, go to [Count events or cart items](https://amplitude.com/docs/analytics/charts/metric-explorer/metric-explorer-choose-measurement#count-events-or-cart-items). ### Average order value by platform, smoothed **Question**: Does average order value differ by platform once daily noise is removed? Set up the chart: - Metric A: `Purchase`, measured as **Average of Property Value** on `order_total`, with **+ Group-by** on the user property `platform`. - On metric A, open the **⋮** menu, choose **Advanced** > **Rolling Window**, and set 7 days. - Optional metric B: `Purchase`, measured as **Event Totals**, with no advanced measurement, to show order volume next to it. **Read the result**: Each point on metric A shows the average order value across the 7 days that end on that day, so weekday and weekend swings flatten out. A gap between platforms that lasts for weeks is a real difference, not noise. Metric B stays raw, because an advanced measurement applies only to the metric you set it on. ### Feature adoption rate as a reusable metric **Question**: What share of paying active users use a feature, and how do you track that ratio across charts? Set up the chart: - Metric A: `Report Exported`, measured as **Unique Users**, with **+ Filter by** `plan` is `paid`. - Metric B: `[Amplitude] Any Active Event`, measured as **Unique Users**, with **+ Filter by** `plan` is `paid`. - Formula C: click **fx** and enter `%:A/B`. Name it `Paid adoption: Report Exported`. - Click **More** above the chart, then **Save as Metric**, and save the metric. **Read the result**: C is the adoption rate as a percentage. **% of Active Users** returns the same ratio when the denominator is every active user. The formula lets you choose the denominator, here paying users only. After you save it, add the metric to other Metric Explorer charts or Data Tables from the metric picker. ## Session analysis ### Sessions, bounce rate, and session duration for new and active users **Question**: Do new users visit as often, bounce as much, and stay as long as the overall active user base? Set up the chart: - Metric A: all sessions (**Total Sessions**). - Metric B: the `[Amplitude] Bounce Rate` metric, from the metric picker. - Metric C: `[Amplitude] Any Active Event`, measured as **Average Session Duration**. - Segment 1: **New** users. Segment 2: **Active** users, which include every user active in the interval. - Visualization: **KPI** to compare the three values for each segment, or **Line chart** for trends. **Read the result**: Each metric keeps its own unit: a count, a percentage, and a duration. If new users bounce more and have shorter sessions than the active baseline, look at the first-visit experience first. User Sessions applies one measurement to the whole chart, so this comparison used to take three charts. ### How many visits include a product view **Question**: What share of visits include at least one product view? Set up the chart: - Metric A: all sessions. - Metric B: `Product Viewed`, measured as **Total Sessions**. This counts sessions with at least one product view. - Formula C: `%:B/A`. **Read the result**: C is the share of visits that reach a product page. To measure visits with deeper browsing, add **+ Filter by** > **\[Amplitude\] Contains Event** > `Product Viewed` **≥** 3 to metric B. Because metric B describes what happened in the visit, you don't need an event that marks a "product visit." ### Bounce rate by entry page **Question**: Which landing pages lose visitors after one page view? Set up the chart: - Metric A: `[Amplitude] Page Viewed`, measured as **Total Sessions**, with **+ Filter by** > **\[Amplitude\] Contains Event** > `[Amplitude] Page Viewed` **=** 1. These are bounced sessions. - Metric B: `[Amplitude] Page Viewed`, measured as **Total Sessions**. - On both A and B, add **+ Group-by** > **\[Amplitude\] First Property Value** > `[Amplitude] Page Path`. - Formula C: `%:A/B`. - To weigh pages by traffic, turn on **Show in results** for metric B. - Visualization: **Bar chart** over the last 30 days. **Read the result**: Each bar is the bounce rate of visits that entered on that page. Sort the breakdown table by metric B to focus on high-traffic pages with a high bounce rate. For how Amplitude defines a bounce, go to [Bounce, entry, and exit](https://amplitude.com/docs/analytics/charts/metric-explorer/metric-explorer-analyze-sessions#bounce-entry-and-exit). ## Events and sessions together ### Add-to-carts compared with sessions over time **Question**: When add-to-carts change, is that because more people visit or because visitors add more? Set up the chart: - Metric A: `Add to Cart`, measured as **Event Totals**. - Metric B: all sessions. - Both are counts, so they share the left axis. If their scales differ, add a right Y-axis and assign metric B to **Right**. **Read the result**: Lines that move together point to traffic. Lines that diverge point to a change in how many items each visit adds. To measure that directly, add formula C: `A/B`, add-to-carts per session. ### Why page views went up **Question**: Page views rose. Is each visit going deeper, or are there more visits? Set up the chart: - Metric A: `Page Viewed`, measured as **Event Totals**. - Metric B: all sessions. - Formula C: `A/B`, pages per session. - Amplitude hides a formula's inputs by default. To plot A and B next to C, turn on **Show in results** for each. **Read the result**: If C stays flat while B rises, you have more visits of the same depth. If C rises while B stays flat, visits are going deeper. If both rise, both are happening, and C shows how much of the change comes from depth. ### Revenue per session by platform **Question**: Which platform earns the most revenue per visit, week over week? Set up the chart: - Metric A: `Purchase`, measured as **Revenue**. - Metric B: all sessions. - Formula C: `A/B`. This is a hybrid formula, because it combines an event metric and a session metric. - Segments: one segment per platform, for example `platform` is `iOS`, `platform` is `Android`, and `platform` is `Web`. - Interval: **Weekly**, which evens out day-of-week swings. **Read the result**: Each line is the weekly revenue per session for one platform. Revenue per session separates visit quality from traffic: a platform can grow revenue only because it gets more visits, while its revenue per session stays flat. ### Conversion rate per session by channel **Question**: Which channels bring visits that convert? Set up the chart: - Metric A: `Purchase`, measured as **Total Sessions**. This counts sessions with at least one purchase. - Metric B: all sessions. - On both A and B, add **+ Group-by** > **\[Amplitude\] First Property Value** > `utm_medium`, so each visit counts under the channel that started it. - Formula C: `%:A/B`. **Read the result**: C is the share of visits from each channel that include a purchase. A user who buys five times in one visit counts once, so a few heavy buyers don't inflate the rate the way a count of purchase events would. ### Visitors who added to cart but didn't buy **Question**: Who added items to a cart and left without buying, and how do you reach them? Set up the chart: - Metric A: `Add to Cart`, measured as **Total Sessions**, with **+ Filter by** > **\[Amplitude\] Contains Event** > `Purchase` **=** 0. - Visualization: **Line chart**, daily. **Read the result**: Each point counts visits with an abandoned cart. Click a point to open Microscope, then click **Create Cohort** to save the users behind it. Use the cohort as a segment in other charts, or target the same users with a guide, survey, or experiment from Microscope's **Add Users to…** menu. For all Microscope actions, go to [Explore a data point with Microscope](https://amplitude.com/docs/analytics/charts/metric-explorer/metric-explorer-microscope).