Skip to main content
In the Analytics section, you can view and analyze the performance of your bots. Is the part of the product that tells you how people are using your chatbots, how many saw them, how many actually talked to them, how many finished, where they gave up and what answers they gave along the way.
Analytics is available on paid plans. On the free and personal plans the section appears locked, with an option to upgrade.

​
Key metrics

Everything in Analytics is built on three simple counts, measured over the period you choose:
  • Views: how many times the chatbot was seen (loaded on a page).
  • Started: how many people actually began a conversation.
  • Completed: how many people reached the end of the conversation.
From those, the system calculates the rates that tell the real story:
  • Started Rate: of everyone who saw the bot, what % started chatting.
  • Completion Rate: of everyone who started, what % finished.
  • Drop-off Rate: of everyone who started, what % quit before finishing.

​
Filters

At the top of the page you have two filters that control everything below them:
  • Time filter: choose the period you’re looking at (e.g. today, last 30 days, year-to-date). Use it to compare periods and spot trends.
  • Variable filter: narrow the analytics down to a single piece of information your bot collects (a “variable”, such as budget, city, or plan). Selecting a variable here is what reveals the Collection metrics section further down the page.
    If your bot doesn’t collect any variables, this filter won’t appear.

​
Summary cards

Shows Total Views, Total Started, and Total Completed for the period you picked, each with its percentage underneath.
Summary Cards

​
Performance chart

A bar-and-line chart showing how those numbers move day by day. The bars show a raw count (e.g. views per day) and the line shows a rate (e.g. completion rate %). You can switch which metric each shows using the two dropdowns. This is where you spot trends — a good day, a bad week, the effect of a marketing campaign.
Performance Chart
Here you can see the total number of views, started sessions and sessions completed.

​
Conversion funnel

A funnel-shaped picture that stacks Views → Started → Completed on top of each other so you can see how many people fall away at each step. A funnel that narrows sharply tells you exactly where you’re bleeding potential customers.
Analytics Conversion Funnel

​
Collection metrics

When you select a variable in the Variable filter at the top of the page, a Collection metrics section appears below the conversion funnel.
Analytics Collection Metrics: Variable
Variable analytics lets you:
  • See the collection rate: of the people who started, how many actually gave you that piece of info?
  • Break results down by answer: e.g. for the question “What’s your budget?”, see how many chose “low / medium / high” and — crucially — the completion rate for each group. Maybe “high budget” people finish 90% of the time and “low budget” people quit early. That’s a real business insight.
  • For numeric answers, see the average value (e.g. average budget entered).
The section includes a table with one row per answer, and these columns: The Completion Rate is color-coded so you can scan it at a glance:
  • 🟢 Green: 75% or higher (this group finishes well).
  • 🟠 Orange: 50 to 74% (mediocre).
  • 🔴 Red: below 50% (this group is dropping off).
Analytics can also show combined numbers across all of your bots at once, not just one bot at a time. Use it for a workspace-wide pulse-check, then drill into an individual bot for detail.

​
Examples

​
1. Business type (button choice)

Suppose your bot asks visitors what kind of business they run and offers three options as buttons — for example Retail, Services, and Manufacturing — saved into a variable such as businessType. Select that variable in the Variable filter, and the Collection metrics section shows you:
  • The collection rate: of everyone who started, what % actually picked a business type. If this is low, the question may be appearing too late or feel intrusive.
  • The percentages table, one row per option, so you can compare how each segment behaves:
At a glance you can see that Retail visitors finish well, while Manufacturing visitors are dropping off — a signal that the flow after that choice needs work for that segment.

​
2. Rating at the end of the conversation (1 to 5)

Now suppose the last step of your bot asks people to rate their experience from 1 to 5, saved into a variable such as rating. Because this is the final question, its collection rate closely mirrors your completion rate — only people who reach the end can leave a rating, so “how many gave a rating” and “how many finished the conversation” are almost the same number. It’s a useful sanity check that the two line up. Since the answers are numeric, the Collection metrics also show the average value — your average rating across all sessions (e.g. 4.3). Watch this number over time: if it drops after you change the flow, your latest version is landing worse with users.