Analytics Guide
Ainisa gives you a full set of analytics screens so you can see exactly how your AI agents are performing — how many conversations they handle, how many turn into leads and bookings, which channels work best, and even the emotional journey your customers go through.
This guide walks through each screen: what it shows, how to read it, and why it matters.
A note on access: Basic Statistics is available to everyone. The rest of the analytics screens are part of a paid subscription. Each screen also respects team permissions, so you can control which team members see which analytics.
A note on filters: Most screens let you pick an assistant and a date range. Paid accounts can look back up to 366 days. Where a screen compares periods, it compares against the previous period of the same length.
Basic Statistics
🔗 https://ainisa.com/business/analytics/basic
What it shows: A high-level overview of your whole team's activity — total chats, total messages, average messages per day, and average messages per chat — plus a trend chart of chats started over time.
How it works: It counts every conversation and message across all your agents for the selected date range and plots the daily trend.
Why it helps: This is your quick pulse check. In one glance you see whether activity is growing or slowing, and how engaged people are (more messages per chat usually means deeper conversations). It's the starting point before you dig into the more detailed screens.
Good to know: Basic Statistics is available on every plan, and it looks at your whole team rather than a single agent.
Advanced Statistics
🔗 https://ainisa.com/business/analytics/advanced
What it shows: The full picture for a single agent (or all agents combined):
- Four key numbers — Conversations, Unique Contacts, Leads, and Bookings — each with a change indicator (↑ / ↓ %) comparing to the previous period, and a small trend line.
- Conversation Outcomes — a simple diagram showing how many of your conversations produced a lead and how many produced a booking.
- A trend chart of conversations started over time.
How it works: It pulls real numbers from your conversations, leads, and bookings, then compares the current period to the one before it. The outcome diagram counts how many distinct conversations reached each result.
Why it helps: This is your main dashboard. The change indicators tell you the direction — is this month better than last? The outcome diagram tells you how effective your agent is at turning conversations into results.
Good to know: Leads and bookings are shown as separate outcomes, not a single funnel — because in most businesses a conversation leads to either a lead or a booking, and one doesn't turn into the other. The outcome branches only appear for the result types your agent actually uses (a restaurant that only captures orders won't see an empty "bookings" branch).
Action Statistics
🔗 https://ainisa.com/business/analytics/action-stats
What it shows: How your agent's actions (the tools it can call — like capturing a lead, checking availability, or making a booking) are being used.
Why it helps: Actions are what turn a conversation into something useful. This screen helps you see which actions fire most often and confirm that the tools you set up are actually being used in real conversations.
Good to know: If an action you expected to be used isn't showing activity, it's a sign the agent isn't triggering it — a cue to check the action's configuration or the agent's instructions.
Lead Tool Analytics
🔗 https://ainisa.com/business/analytics/lead-tool-analytics
What it shows: A detailed log of every time your agent's lead-generation tool ran — with the captured fields (name, phone, order details, and whatever else you configured), the status of each execution, and the date. Up top: total runs, how many succeeded, how many failed, and the success rate.
How it works: Pick an assistant and one of its lead actions. The screen lists each tool execution, newest first, with search and page-by-page navigation so you can move through large volumes.
Why it helps: This is where you verify the tool is working correctly. If leads aren't reaching your CRM or sheet, the success/failed counts tell you immediately whether the problem is on the agent's side. You can also search the raw captured data.
Good to know: This screen is about the tool's behaviour (did the call succeed?), which is different from the Lead Analytics screen below (what happened to the lead afterwards).
Booking Tool Analytics
🔗 https://ainisa.com/business/analytics/booking-tool-analytics
What it shows: A detailed log of every time your agent's booking tool ran, broken down by operation. Up top: how many bookings were booked, rescheduled, and cancelled, plus a net active figure (booked minus cancelled).
How it works: The booking tool does several things — checking availability, booking, rescheduling, cancelling. This screen counts only the real booking events (book / reschedule / cancel) for the KPIs, and lists every execution with the guest details, operation, status, and date. Search and pagination included.
Why it helps: It shows not just how many bookings were made, but how many were cancelled — the number that's easy to miss but matters most. A high cancellation count next to your bookings is an early warning sign.
Good to know: Operations like "check availability" and "list bookings" are noise for the headline numbers, so they don't count toward the KPIs — only real booking events do. The net active figure is an approximation; for exact booking status, use the Booking Analytics screen below.
Lead Analytics
🔗 https://ainisa.com/business/analytics/lead-analytics
What it shows: The real status of your leads — how many are New, Seen, Contacted, Confirmed, and Completed, shown as a funnel, with Cancelled tracked separately. Up top: total leads, completed, confirmed, and a conversion rate.
How it works: It reads the actual status of each lead (the status you and your team update as you work them) and groups them. The conversion rate is Completed ÷ Total.
Why it helps: This answers "where do my leads stand?" The funnel shows where leads pile up or drop off. If lots of leads are stuck at "Contacted" and few reach "Completed", that's a follow-up problem you can act on.
Good to know: Conversion counts only Completed leads — the final, real outcome. "Confirmed" is shown as its own number (a pipeline indicator) but isn't counted as converted, because a confirmed lead can still be completed or cancelled later.
Booking Analytics
🔗 https://ainisa.com/business/analytics/booking-analytics
What it shows: The real status of your bookings — Pending, Confirmed, and Completed as a funnel, with Cancelled tracked separately. Up top: total bookings, completed, confirmed, and a conversion rate.
How it works: It reads the true status of each booking record and groups it. Conversion is Completed ÷ Total.
Why it helps: This tells you how many of your bookings actually happened versus fell through. The Cancelled figure, shown separately, is the one to watch — a healthy booking flow keeps cancellations low.
Good to know: As with leads, conversion counts only Completed bookings. "Confirmed" is a pipeline signal (booked but not yet fulfilled), not a converted result.
Channel Analytics
🔗 https://ainisa.com/business/analytics/channel-analytics
What it shows: How each channel performs — Website, Telegram, WhatsApp, Instagram, Facebook, and so on. For every channel you see conversations, leads, bookings, and the conversion rates, both as a comparison chart and a detailed table.
How it works: It groups your activity by the integration channel each conversation came through, then lines up the results side by side.
Why it helps: This answers the question every business owner asks: "which channel actually brings me results?" A channel might bring lots of conversations but few bookings, while another brings fewer conversations but converts far better. That tells you where to focus your time and budget.
Good to know: Only conversations tied to a known channel are counted, so channel totals can be slightly lower than your overall totals. That's expected — a result whose channel can't be determined simply isn't attributed to one.
State Analytics
🔗 https://ainisa.com/business/analytics/state-analytics
What it shows: The emotional and intent journey of your customers, as detected by your AI. As conversations unfold, the agent recognises conversation states — objection, high intent, consideration, confusion, escalation, retention, and more. This screen shows how many conversations reached each state, colour-coded by meaning, plus a heatmap of how customers move from one state to another.
How it works: Your agent runs in a mode (Sales, Support, Booking, Receptionist), and each mode has its own set of possible states. As the AI detects a shift in the customer's situation, it records the transition. This screen groups those detections for the agent you select.
Why it helps: This is analytics no basic tool gives you. It shows where customers hesitate and where they commit. Lots of "objection" means price or trust concerns. Lots of "escalation" means the AI is hitting its limits. Lots of "high intent" means your agent is doing its job. The colours make it readable at a glance — green is good, amber needs attention, red is a churn risk.
Good to know: The states shown depend on the agent's current mode, so the screen asks you to pick a specific agent. The transition heatmap only appears when there are enough transitions to be meaningful — a handful of conversations won't produce a useful map.
AI Usage Statistics
🔗 https://ainisa.com/business/analytics/usage
What it shows: How much AI your agents are consuming — the usage that drives your costs (things like tokens processed and, where relevant, audio usage).
Why it helps: AI has a real cost, and this screen keeps it visible. It helps you understand which activity is driving spend and plan accordingly, so there are no surprises.
Good to know: Usage is the raw consumption behind your agents' work — a useful companion to the results screens, so you can weigh outcomes against what they cost.
Which Screen Should I Use?
| I want to… | Go to |
|---|---|
| Get a quick pulse of overall activity | Basic Statistics |
| See my main dashboard with trends and outcomes | Advanced Statistics |
| Check whether my tools are firing correctly | Lead Tool / Booking Tool Analytics |
| See where my leads and bookings stand | Lead / Booking Analytics |
| Find out which channel brings the best results | Channel Analytics |
| Understand where customers hesitate or commit | State Analytics |
| Keep an eye on AI cost | AI Usage Statistics |
Most businesses start with Advanced Statistics for the daily picture, then use Channel and State Analytics to understand why the numbers look the way they do.