Skip to main content

Reading Your Upsell Analytics

Understand the four upsell metrics, how revenue attribution works, and how to compare performance over time.

Written by Tom Nipravsky

Every upsell tracks its own performance, and the Upsells list page rolls those numbers up into a store-wide view. This article walks through the four metrics, explains how revenue is counted, and offers a few questions you can ask the data to decide what to keep, pause, or change.

The list page at a glance

The Upsells list page has four metric tiles at the top:

  • Upsell Revenue

  • Views

  • Conversion Rate

  • AOV Lift

Below the tiles is a daily revenue chart showing upsell revenue over time. Below the chart is a table listing every upsell — name, type, discount, status, plus per-upsell views, conversions, and revenue.

Each metric tile shows the percentage change versus the previous period of the same length as a small badge:

  • A green up-arrow badge means the metric improved (or revenue/views grew, or conversion rate climbed)

  • A red down-arrow badge means the metric dropped

  • A neutral badge appears when there's no change

For example, on a 30-day view the comparison is to the prior 30 days. On a 7-day view it's to the prior 7 days.

What each metric means

Upsell Revenue

The total gross revenue attributable to upsells over the selected date range, in your shop currency. Counted at the gross price of the upsold items — discounts you offered on the upsell are reflected, but tax and shipping are not.

Revenue is added to the metric only when the order is paid. See How attribution works below.

Views

The number of times an upsell was rendered to a customer in the selected range. One view is recorded each time an upsell section appears in front of a buyer — on the Checkout cart, the Thank You page, or the Order Status page.

If the same customer reloads the Order Status page three times, that's three views.

Conversion Rate

The percentage of views that turned into a paid upsell — i.e. Conversions ÷ Views × 100.

A few things to keep in mind when reading this number:

  • The denominator is views, not unique customers. Reloads inflate views and slightly suppress the conversion rate.

  • Conversions are counted at payment, not at the moment the customer clicks Add. An add that's never paid for doesn't count.

  • Industry "good" benchmarks vary wildly by store and product type. Treat the number as a baseline to compare against your own future periods rather than an external benchmark.

AOV Lift

The incremental average order value driven by upsells over the selected range. It tells you, on average, how much more each order is worth thanks to upsells.

A positive AOV Lift means upsells are pulling order values up — typically what you want. A flat or negative lift suggests the discount you're offering may be too steep, or the upsells are cannibalising items the customer would have bought anyway.

The daily revenue chart

Below the tiles is a bar chart showing Upsell Revenue by day over the selected range. It's most useful for two questions:

  • "Is performance steady or trending up/down?" — a clean upward slope is what you want; flat or downward suggests the upsells are stale.

  • "Was there a specific day something changed?" — sudden jumps or drops often correlate with a discount change, a new upsell going live, or a marketing campaign that brought different traffic.

Per-upsell analytics

When you open an individual upsell to edit it, an analytics panel appears alongside the form showing the same four metrics — but scoped to just that upsell over the selected range. Use this when you want to know how a specific upsell is performing without the noise of the rest of the dashboard.

The list page table also shows per-upsell views, conversions, and revenue at a glance — useful for spotting outliers without clicking into each upsell.

How attribution works

The most important thing to understand about upsell analytics is when revenue is counted.

  • A view is recorded the moment an upsell renders in front of a customer — checkout cart, Thank You page, or Order Status page.

  • A conversion (and revenue) is recorded only when the order is paid.

For Checkout upsells the customer pays in the same flow, so the gap between "view" and "conversion" is seconds.

For Order-editing upsells there can be a gap. A customer can add an upsell on the Thank You page, but if they never complete the additional payment for the new total, the conversion is not counted. The same applies on the Order Status page: an add only counts once the resulting balance is settled.

This is intentional. The conversion metric reflects money you've actually collected, not click-through optimism. It also means:

  • An accepted upsell may take a little time to appear in your conversions count if the customer doesn't pay immediately.

  • Abandoned upsells (customer adds, never pays) don't pollute your numbers.

  • A customer who later cancels the order does not retroactively reduce the conversion count.

Date range and period comparison

The list page has a date range picker at the top with the usual presets — Today, Last 7 days, Last 30 days, plus a Custom range picker.

Whatever range you pick, the delta badges on the metric tiles compare to the previous period of the same length:

  • Last 7 days → compared to the 7 days before that

  • Last 30 days → compared to the 30 days before that

  • A custom 14-day range → compared to the 14 days immediately before it

This makes it easy to ask "are we doing better than last week?" without doing the math yourself.

Tip: When you launch a new upsell or change a discount, give it at least 7–14 days before reading the analytics seriously. A few hundred views is the bare minimum for the conversion rate to settle.

Reading the numbers — questions to ask

Once you have a couple of weeks of data, the dashboard becomes useful for decisions. Some questions worth running through:

  • Which upsell is converting best? Sort the per-upsell table by conversion rate. The top performer is your benchmark — try to apply what's working there to weaker upsells (similar product source, similar audience).

  • Are my upsells lifting AOV? If AOV Lift is flat or negative, your discounts may be too generous, or the items being upsold are cannibalising — consider tightening filters or shrinking the discount.

  • Which upsells should I pause? If an upsell has a meaningful sample of views (say 200+) and the conversion rate is well below your other upsells, pause it and try a different product source or filter.

  • Is the Thank You page or Order Status page driving more revenue? Run two Order-editing upsells with Upsell ID pinned to each surface (see Upsells in the Order-Editing Flow) and compare per-upsell numbers.

  • Did revenue jump after I made a change? Use the daily chart to align the visible change with the date you tweaked the configuration.

What's next

Did this answer your question?