Analytics

Know what your agents are doing for the business

Analytics turns every conversation into numbers you can act on. See volumes and response times, where conversations drop off in a flow, what the AI costs, which agent books the most, and how your website chat performs, filtered by agent, inbox and date range.

An analytics dashboard with KPI tiles, a conversations trend line and bookings by agentAnalytics · last 30 daysConversations12,480+18%Answered by AI94%+3 ptsMedian first reply38 s−12 sBookings1,126+24%Conversations per dayTue 24542 convosBookings by agentAva · Studio447Max · Dental333Kai · Surf213Nia · Realty146
10analytics views
Per stepflow drop-off
Per runAI cost
LiveA/B results
Deep dive

A closer look at Analytics

01 · Flow

See exactly where conversations drop off

The flow view overlays real traffic on your job flow: how many conversations reached each step, which exits they took and where they stalled. The conversation map shows the most common paths, so you know which step to fix first.

  • Conversations per step and per exit
  • Drop-off highlighted on the flow
  • Conversation map of common paths
  • Filter by agent, inbox and dates
A flow drop-off view where the conversation stream narrows from 1,000 at Start to 351 booked, with the biggest drop between Qualify and Offer highlighted−18018% stalled−21026% stalled−20834% stalled−5113% stalledStart1,000100%Qualify82082%Offer61061%Book40240.2%Booked35135.1%Biggest drop-off: Qualify → Offer (−26%)Open this step in Test & Debug to see why conversations stall here
02 · Speed

Response times your customers will notice

Track first-response and reply times across channels, with a breakdown by inbox and agent. Spot slow providers, busy hours and business-hours waits at a glance.

  • First-response and reply time distributions
  • Breakdown by inbox, channel and agent
  • Volume over time
  • Overview KPIs at the top
A histogram of first-response times with the median and 90th percentile marked, and a breakdown by channelFirst response timemedian 38 sp90 2m 10s0 s5 minBy channel (median)Web chat9 sSMS22 sWhatsApp31 sEmail1m 40s
03 · Outcomes

Leads, bookings and cost per result

The leads view shows who became a lead and who booked, attributed to the agent that did it. The AI view shows token spend and cost, so you can see cost per booking, not just cost per message.

  • Leads and bookings per agent
  • AI cost and tokens over time
  • Quote desk and web chat views
  • A/B results alongside everything else
Leads and bookings per agent, lead sources by channel, and cost per bookingLeads → bookings by agentAva312 leads188 bookedMax240 leads131 bookedKai176 leads72 bookedLeads by channel728leadsSMS 42%Web chat 24%Email 20%Instagram 14%$0.41AI cost per booking54%lead → booking rate3.2messages to book
Under the hood

The details

Overview

Volume, reply rate and headline KPIs.

Response

How fast agents and teammates respond.

Breakdown

Compare inboxes, channels and agents.

AI

Tokens, cost and model usage.

Agents & flow

Per-agent performance and per-step drop-off.

Leads, web chat, quotes

Bookings, widget conversations and quote desk throughput.

Use cases

Who it's for

Weekly reviews

Spot the step with the biggest drop-off and fix it in Test & Debug.

Client reporting

Show agencies' clients bookings made by their agent.

Budgeting

Track AI spend against leads and bookings.

FAQ

Analytics: common questions

Can I filter analytics by agent?

Yes, and by inbox and date range.

Are test conversations included?

No. Analytics counts real conversations.

Can I see AI costs?

Yes. Each run records tokens and cost, rolled up in the AI view.

Where are A/B results?

In their own Analytics tab and on the A/B node in the builder.

Get started

Put it to work on your own inbox

Connect an inbox, describe the job, test it in a sandbox and go live with exactly the autonomy you're comfortable with.