How to understand chatbot analytics and KPIs
Turn dashboard metrics into defensible product, content and conversion decisions.
A KPI becomes useful only when its definition, timeframe and comparison group are clear. WebChatAgent exposes eight headline metrics plus charts and tables for deeper review.
Do not treat message volume as success by itself. Pair activity with outcomes, unanswered questions and the visitor journey.
What you will have at the end
- A documented KPI glossary
- A clean date and chatbot comparison
- One evidence-based optimization hypothesis
Before you start
- At least one chatbot with activity
- A business outcome such as lead or booking
- A comparison period with similar traffic
Activity, quality and outcome are different layers
Sessions and messages describe activity. RAG messages and unanswered questions indicate knowledge usage and gaps. Leads, bookings and conversion describe outcomes.
Filters define the denominator. A conversion rate across all bots and a single campaign bot answer different questions.
01–05
Set it up step by step
Open Analytics and set the cohort
Choose chatbot and date range before reading numbers.
Open Dashboard → Analytics, select one chatbot or all, then set a date range aligned with the business question. Avoid comparing a full month with a partial week.
Review sessions and message activity
Understand conversation volume and depth.
Read Sessions, User Messages and Assistant Messages together. A rise in messages per session can mean deeper engagement or more difficulty; inspect conversations before assigning meaning.
Review RAG and unanswered metrics
Measure use of knowledge and missing coverage.
Compare RAG Messages with Unanswered Questions. A low unanswered count is useful only when difficult real questions are still being asked and correctly classified.
Read leads, bookings and conversion
Tie outcomes to the intended chatbot role.
For a sales bot, Leads and Conversion may matter most; for scheduling, Bookings matters. Verify how the dashboard defines conversion and use the same definition across comparisons.
Compare chatbots and inspect records
Move from aggregate signal to concrete evidence.
Compare bots only when traffic source and purpose are comparable. Open related conversations, questions, leads or bookings to identify the content or flow behind the KPI change.
Example & result
See the practical test and its result
Every tutorial includes a fixed input, the expected outcome and a transparent record of what was actually verified locally.
Practical example: How to understand chatbot analytics and KPIs
This exact scenario was completed with the temporary tutorial account.
Exact test input
Filter Analytics to “Tutorial Lab” and compare the live signals recorded on 17 July 2026.
Expected result
The chatbot filter, KPI cards and daily row report the same visitor activity.
What was actually verified
The aggregated result showed 2 sessions, 5 user messages, 5 assistant messages and 5 RAG messages in both KPI cards and the daily table.
Tips & tricks
Make the setup reliable
Test with realistic examples, record your baseline and change one setting at a time. That makes real improvements visible.
Add annotations outside the dashboard
Record launches, campaigns, model changes and source re-indexing so metric shifts have context.
Prefer rates plus counts
A high conversion rate from three sessions is not equivalent to the same rate from three thousand.
When something does not work
Troubleshooting
Check status, permissions and test data systematically before changing the model or prompt.
The expected option is missing
Confirm the account plan, feature permissions and selected chatbot. Paid or beta features can be hidden when prerequisites are not met.
The test result is inconsistent
Reset the test conversation, keep the input identical and change one setting at a time so the cause remains measurable.
Ready for a production-style test
Choose one KPI movement, inspect at least ten underlying records and make one reversible change. Measure again over a comparable window.