Onsite messaging performance report

The performance report shows how your onsite messages perform: how often visitors see them, how many interact, how many sign up or share, and the purchases that follow. You can compare messages side by side, including the variants in an A/B test, and follow the results day by day in the charts. This article shows you how to open the report, choose what it shows, and read each chart and metric, including how every number is counted.

Open the performance report

In Customer Engagement, go to Onsite > Performance.

The same report is also in the Onsite Manager, under Insight Reports > Performance. Both show the same data, whichever editor you build your messages in.

Choose what the report shows

Use the filters at the top of the page, and then select Filter to update the report.

The top of the performance report, with the State, Select site, date range, and Deleted filters, the Filter and Export buttons, and the charts

Filter What it does
State Shows only messages in the states you choose: Live, Stand by, Expired, or Draft. Leave it empty to show messages in every state.
Select site Shows only messages on the site you choose.
Date range The period to report on. Choose a start date and an end date in the calendar; both days are included. By default, the report shows the last seven days and today.
Deleted Includes messages that have been deleted, so you can still see their results.

To download the report, select Export. The button shows when there's data for the period.

The charts

The charts show the messages and period you've filtered on, one point per day. The two charts at the top are always shown. Use the arrows next to 5 Graphs to scroll through the others.

Chart What it shows
Widget impressions How many times your messages were shown maximized, per day.
Contact database growth A running total, over the period, of New contacts/members from onsite messaging and New reachable contacts from onsite messaging. The line for the last day matches New contacts and New subscribers in the table.
Interaction rate (%) The interaction rate per day. See Interaction rate below.
Traffic to site Visits to your website generated by your messages: Return visits (through links in your confirmation and reminder emails), Friend visit, and Friend return visit.
Conversions Purchases linked to your messages, made by visitors (Conversions) and by friends in refer-a-friend campaigns (Friend conversions).
Revenue (aggregated) The revenue from those purchases.
Average order value The average value of those purchases.

The table

The table has one row per message, with the message name and the site under it, and the message's state. Select a name to open the message in the editor. The first row shows the totals for all messages in the table.

The performance report table with the Show all, Referrals, Signup, and Onsite promotion tabs, the totals row, and one row per message

Every message has its own row, so in an A/B test, each variant has its own row too. You can compare them directly, including an empty variant used as a control group. To learn how one message is chosen when several use the same script tag, see How a message is selected.

The table scrolls sideways. Select a column heading to sort by it, and point to the ⓘ icon next to a heading to see a short description.

The tabs above the table choose which columns are shown. Every tab lists the same messages.

  • Show all: every column.
  • Referrals: the columns for refer-a-friend campaigns, including the friend columns.
  • Signup: the columns for signups, such as New contacts, New subscribers, and New Data Points.
  • Onsite promotion: the columns for promotions, such as sessions, impressions, and conversions.

The metrics

All rates are percentages, rounded to two decimals. A rate shows 0% when there's nothing to divide by, for example when no rewards have been issued.

Visits and impressions

Metric What it counts
Link clicks / QR code visits Visits to your website through the message's own link or QR code.
Sessions The number of times a page that calls the message's script tag was loaded while the message was active. Previews aren't counted. Each page load counts, so one visitor who browses five pages counts five times, and a visitor on two devices counts on both.
Minimized impressions The number of times the message was shown minimized.
Maximized impressions The number of times the message was shown maximized, for example when it opens automatically or a visitor selects the minimized message.
Interaction rate Maximized impressions divided by Sessions. It's most useful for messages that start minimized, since it shows how often visitors open them. It can go over 100% if a message is maximized more than once on the same page.

Completions and sharing

Metric What it counts
Completed In each message's row, the number of times visitors selected the message's Close button. In the totals row, the number of times visitors selected Finish. Because they count different buttons, the total can differ from the sum of the rows.
Completion rate Completed divided by Maximized impressions.
Started conversations Refer-a-friend campaigns only. The number of times visitors shared your offer by selecting a sharing channel, such as WhatsApp or a link. One visitor can share through several channels.
Started conversations rate Started conversations divided by Maximized impressions.
Users engaged Refer-a-friend campaigns only. The number of visitors who shared your offer at least once. Each visitor counts once, however many channels they used.
Engagement rate Users engaged divided by Maximized impressions.
For signups, Started conversations, Users engaged, and Engagement rate are always 0, since they only count sharing. To see how many visitors signed up, use New contacts and New subscribers.

Rewards, visits, and purchases

These columns are for the visitor who sees your message. In refer-a-friend campaigns, that's the visitor who shares, and their friends have their own columns.

Metric What it counts
Rewards issued The number of rewards, such as discount codes, given to visitors. Friend rewards aren't included.
Return sessions to site (the first column with this name) Visits to your website through the links in the confirmation and reminder emails the visitor gets. Friends' visits aren't included.
Return sessions to site (the second column with this name) Conversions: the number of visitors who made a purchase within the attribution time, whether or not they used the reward. Friends aren't included. See How purchases are linked to a message.
Conversion rate (the first column with this name) Visitor conversions divided by Rewards issued. It shows how many of the visitors who got a reward made a purchase.
Revenue The value of the visitors' purchases. Friends' purchases aren't included.
Average order value Revenue divided by the number of visitor conversions.

Friends in refer-a-friend campaigns

Metric What it counts
Friend sessions to site Visits to your website by friends, through the referral link and through the links in the friend emails.
Friend sessions per ambassador Friend sessions to site divided by Users engaged. It shows how many friend visits each visitor who shared brought in, on average. An ambassador is the visitor who shares.
Friend rewards issued The number of rewards given to friends.
Friend conversions The number of friends who made a purchase within the attribution time, whether or not they used their reward.
Friend conversion rate Friend conversions divided by Friend rewards issued.
Friend revenue The value of the friends' purchases.
Friend average order value Friend revenue divided by Friend conversions.
Friend reward issue rate Friend rewards issued, divided by the number of friends' visits through the referral link.

Totals for visitors and friends

Metric What it counts
Total clicks All visits to your website generated by the message: visits through the email links, friends' visits, and visits through the message's link or QR code.
Total rewards issued Rewards issued plus Friend rewards issued.
New contacts The number of new contacts or members created in Customer Engagement, by visitors and friends.
New subscribers New contacts plus existing contacts who became reachable, for example by giving consent to emails, by visitors and friends.
New Data Points The number of new pieces of personal data saved on contacts, such as a phone number or a birth date, by visitors and friends.
Total conversions Visitor conversions plus Friend conversions.
Conversion rate (the second column with this name) Total conversions divided by Total rewards issued.
Total revenue Revenue plus Friend revenue.
Currency The currency the revenue is shown in.

How purchases are linked to a message

A purchase counts as a conversion for a message when the visitor, or their friend, buys something within the attribution time. It counts whether or not they used the reward.

  • The attribution time is set for the whole site in Default attribution time (days) and, for friends, Default friend attribution time (days). See Onsite messaging site settings.
  • A single message can use its own attribution time with Specific attribution time (days).
  • Purchases are only registered when your checkout page calls the onsite messaging script, with the order details. Ask your developer if no conversions show up.

Tips for reading the results

  • Compare rates, not totals. Messages run for different lengths of time and on different pages, so compare Interaction rate, Completion rate, and Conversion rate.
  • Check that a minimized message gets opened. A low Interaction rate on a message that starts minimized means few visitors open it. Try a different text or placement, or let the message open automatically after a delay.
  • Measure signups with contacts. For a signup, New contacts and New subscribers show how many visitors signed up, better than Completed.
  • Compare A/B variants over the same period. Choose the same dates for all variants, and give the test enough time before you decide.
  • Expect differences from your web analytics. Tools such as Google Analytics count sessions, visits, and purchases in their own way, so their numbers won't match the report exactly.

Was this article helpful?

/