Why they leave
Read what visitors said before leaving without booking, see what your follow-ups recovered, and get evidence-based suggestions for your qualification rules.
What this tab is for
Analytics tells you how many visitors book. Why they leave tells you why the others did not, in their own words. Every conversation that went quiet and every form abandoned after the first answers is a first-hand account of what stopped a visitor. Nudge reads them and turns them into a dated report.
Open Analytics, then the Why they leave tab. The period and form filters of the Analytics bar do not apply here: this tab has its own.
Reading reports requires the analytics permission. Generating a new report requires the forms management permission.
Generate a report
Choose the period
Under New report, pick 7 days, 30 days or 90 days.
Choose the form
All forms or a single form.
Click Analyse
Nudge's AI reads the conversations that did not end in a booking and the abandoned forms of the period. It takes up to a minute.
The report needs at least 5 abandonments in the period. With fewer, Nudge tells you how many it found and generates nothing.
Read the report
Each report shows its period and its sample ("12 conversation(s) and 4 abandoned form(s)"), so you always know how much evidence a conclusion rests on. Then:
- a short summary,
- up to five causes, each with the number of conversations it covers and verbatim quotes from visitors,
- The question that blocks, when one question of your form makes visitors leave,
- Change: a concrete suggestion for each cause.
If nothing stands out, the report says "No clear cause stands out in this sample." Reports are kept, newest first, so you can compare this month with last month.
Privacy
Reports work on what visitors said, not on who said it. Names, emails and phone numbers are removed before the conversations are analysed and are never stored in a report.
Follow-ups
The Follow-ups card shows what the follow-up emails of the Nurturing Agent produced over the selected period:
- Sequences sent,
- Conversations resumed (visitors who came back to the chat from an email),
- Bookings recovered,
- with the number of emails sent, opened and clicked.
To set up these emails, see Workflows.
Suggested qualification changes
Nudge compares the answers of visitors who booked with those who did not, over 90 days of conversations, and proposes changes to a form's qualification rules.
- Choose a form, then click Analyze rules.
- Each proposal reads like "Proposal: exclude "Under 500 €" from qualification", with its sample ("40 conversations, 2 meetings in this group").
- Click View examples and change to read the conversations behind it, then Apply or Dismiss. An applied change can be undone with Revert change.
A proposal needs at least 20 comparable answers and 10 per group. Nothing changes automatically. Keep in mind that a visitor who did not book is not necessarily a poor prospect: read the examples before applying. See Qualification and scoring.
What they ask for
The What they ask for card counts, over the period, what visitors raised in conversations that have a summary:
- Pain points,
- Objections,
- Topics raised,
each with how many conversations mention it and how many of those led to a meeting. This card involves no AI call: it counts what is already in the conversation summaries.
Also available through MCP
An AI assistant connected through MCP can generate an abandonment report, list past reports, read the follow-up statistics, read the demand map (pain points, objections, topics) and list the pages that drive bookings. See API & MCP.