Outbound Virtual Agents Reporting and ROI Measurement Guide

Overview

Outbound Virtual Agents is the AI Lead-follow-up capability in the Growth section of ServiceTitan. Once your campaigns have been live for a few weeks, the question stops being "is it running?" and becomes "is it paying for itself?" This guide is written for the manager or owner who has to answer that second question — it explains the performance data Outbound Virtual Agents produces, defines each KPI in plain terms, and walks through a defensible way to turn those numbers into a return-on-investment figure you can stand behind.

This is a post-go-live guide. It assumes your campaigns are active and have accumulated enough conversations, bookings, and completed jobs to be worth measuring. If you are still setting up, read Outbound virtual agents overview in ServiceTitan Max: Configuration and management first, and have your team work through Utilizing Speed to Lead as a CSR in ServiceTitan Max for the day-to-day operator view.

Because you are working in ServiceTitan Max, the numbers you review here are the visible result of all features working together, not the AI agent alone. Adaptive Capacity supplies the Appointment times the agent quotes, Contact Center is the engine the Voice Agent and SMS Agent run on, and Marketing supplies the attribution and audience data that decides which leads enter your campaigns. When a number moves, the cause is often upstream of the agent — a point this guide returns to repeatedly, because it is the difference between reacting to a symptom and fixing a cause.

Who this guide is for

Managers and Owners measuring the return on Outbound Virtual Agents
Anyone signing off on whether to expand, adjust, or hold campaign investment

For the click-by-click operator workflow — where each metric lives on screen and how CSRs take over conversations — see the Insights and Speed to Lead CSR articles cross-referenced at the end.

What you need to review this data

Outbound Virtual Agents and Speed to Lead are included in ServiceTitan Max. If your Growth section is not yet active, your administrator should reach out to your PPS.
The following permissions let you and your team see the underlying conversations and revenue. Contact your account administrator to confirm them:
- View Customer Chats — required to open the Interactions page and read conversation transcripts
- Send Message — required to use Takeover on an in-progress interaction
- Listen to Call Recording — required to view Call transcripts and recordings in the interaction detail drawer

Where the numbers live

Outbound Virtual Agents reporting has two homes, and knowing which one answers which question saves time.

The Insights page (Growth > Insights) is your primary reporting surface. It shows how leads move from first contact through to a booked, completed Job — across all active goals and broken down per goal. This is where response rate, conversions, jobs completed, and revenue all live, and it is where the day-to-day and week-to-week measurement happens.

The Interactions page (Growth > Interactions) is the conversation-level log — every SMS and voice exchange, with an AI-generated summary, status, and full transcript. Managers use it less for measurement and more for spot-checking quality and understanding why the Insights numbers look the way they do.

A few things to know before you read any number:

Insights data is filtered by time period. Use the Time Period dropdown at the top of the page to set the window. The default is the last 30 days. Apply a consistent window whenever you compare across periods.
Top-level metrics span all active goals. Click into a goal card (for example, Speed to Lead) to see that goal on its own.
Jobs and revenue reflect only work originated by Outbound Virtual Agents. Bookings the AI agent creates are tagged for reporting, which is what makes clean attribution possible. This tagging is also what lets those same jobs flow into Marketing Revenue Attribution.
Exporting interaction data is not yet available. Interaction logs and transcripts are reviewed directly in the Interactions page. Insights metrics, however, can be pulled into ServiceTitan reports through the tagged bookings.

KPI definitions

The metrics below are grouped the way they appear in Insights: the top-level summary cards, the goal-level summary, the Conversion Funnel, and the three detailed metric sections. Wherever this guide uses a plain-language term (for example, "response rate" or "jobs completed"), the exact in-product label is given so the number you read here matches the number on your screen.

Top-level summary metrics

These four cards sit at the top of the Insights page and summarize every active goal for the selected time period. Each card also shows a trend indicator comparing the current window to the previous equivalent one.

Leads — the total number of leads contacted by Outbound Virtual Agents in the selected period.
Avg Response Rate (your "response rate") — the percentage of leads who responded to at least one outreach attempt. This is the clearest early signal of whether your messaging and cadence are landing.
Conversions — the number of leads that resulted in a booked job.
Avg Conversion Rate — the percentage of leads that converted to a booked job.

Goal-level summary

Below the top-level cards, each goal (such as Speed to Lead) has its own summary card showing:

Conversion Rate for that goal
Response Rate for that goal
Total Revenue generated by bookings from that goal

Clicking the arrow on a goal card opens its detailed Insights view, which contains the Conversion Funnel and the three metric sections below.

The Conversion Funnel

The Conversion Funnel is the single most useful diagnostic view for a manager. It shows the count of records at each stage of the outreach sequence, so you can see exactly where leads fall away. Each bar also shows the percentage change versus the previous period.

Leads — total leads that entered the campaign
Conversations Started — leads the agent initiated contact with
Responses — leads who replied to at least one message
Conversions — leads who converted (job booked)
Jobs Booked — confirmed booked jobs
Jobs Completed (your "jobs completed") — jobs that have been completed

Reading the funnel is a matter of finding the biggest drop between two adjacent stages. A large fall between Leads and Conversations Started usually means an audience or attribution issue in Marketing — the leads exist but are not matching your campaign criteria. A large fall between Responses and Conversions usually means Adaptive Capacity did not have availability to quote, or the agent's job-handling configuration in Contact Center needs adjustment. The funnel tells you which system to look at, not just that something is off.

Conversion metrics

Conversion Rate — percentage of leads that became booked jobs
Conversions — total number of converted leads
Leads — total leads in the period
Median Time to Convert — the median time from first contact to booking

Jobs & Revenue metrics

This section is the heart of any ROI conversation, because it is where booked work becomes measured revenue.

Jobs Booked — total jobs booked by the AI agent
Jobs Completed — total completed jobs originating from Outbound Virtual Agents
Completed Job Revenue (your "revenue attributed") — total revenue from completed jobs
Average Ticket — average revenue per completed job

A note on the difference between booked and completed: a booked job is revenue you expect; a completed job is revenue you earned. ROI should be built on Completed Job Revenue, because a booked job that cancels or reschedules out of the window never becomes cash. The gap between Jobs Booked and Jobs Completed is itself worth watching — a widening gap points to booked work that is not making it to completion.

Performance metrics

Median Speed to Lead Time — the median time between a lead entering ServiceTitan and the agent's first outreach. For a Speed to Lead campaign this is the metric that most directly reflects the promise of the feature: contacting a lead in seconds, before a competitor does.
Opt Out Rate — the percentage of leads who opted out of further messages. A rising opt-out rate is an early warning that message frequency or tone is wearing on your audience, and it is worth acting on before it drags response rate down.

How the numbers connect to the rest of ServiceTitan Max

The reason a manager should resist reading any single metric in isolation is that almost every number on the Insights page has an upstream owner. Understanding those links turns "the number dropped" into "here is what to fix."

Marketing supplies the attribution and audience data that decides which leads enter your campaigns. If Leads in the funnel is lower than expected, the cause is usually upstream in Marketing's audience configuration or lead-source tracking, not in the agent.
Contact Center is the engine the Voice Agent and SMS Agent run on. The persona, transcript, skills, and escalation rules configured there shape how the agent talks to leads and how quickly conversations close, so Response Rate and Median Time to Convert are downstream of those settings.
Adaptive Capacity provides the Real-time availability the agent quotes. Conversion Rate depends heavily on whether availability exists at the moment the agent is talking to a lead. A high Response Rate paired with a low Conversion Rate points at Adaptive Capacity first.
Scheduling booking links embedded in agent messages give leads a self-serve path; conversions that happen that way still flow into the Conversions and Jobs Booked counts.
ServiceTitan Core is where booked jobs land and where Jobs Booked, Jobs Completed, Completed Job Revenue, and Average Ticket are sourced from.

The practical takeaway for a manager reviewing performance: when a metric moves, trace it to the system that owns it. A drop in Median Speed to Lead Time often points to a Marketing attribution gap or a Voice Agent setting in Contact Center. A drop in Conversion Rate with a steady Response Rate points to Adaptive Capacity availability.

Pulling your performance data

Insights gives you the live, on-screen view. When you need the numbers in a report you can save, schedule, or bring to a review meeting, the tagged bookings the AI agent creates can be pulled through ServiceTitan reports. A few practices keep the data trustworthy:

Set a consistent date range. Use the same window for every comparison. A 90-day window is the standard first checkpoint for ServiceTitan Max because it is the first point at which the AI-driven capabilities have enough live data to be measured meaningfully.
Confirm Business Unit filters. If your business units are not mapped to your trades or divisions, your numbers will be incomplete or wrong.
Confirm lead-source attribution. Verify that jobs booked by the AI agent are tagged with the correct source, so Marketing Revenue Attribution captures them. Attribution gaps here are the most common reason ROI looks lower than it actually is.
Assign one owner. Reviews stall when the data lives with one person and the follow-ups with another. Put one manager or admin in charge of the pull, the findings, and the actions.

For the full report-by-report walkthrough of how to extract this data, see Pulling Your 90-Day Max Performance Data in ServiceTitan Max.

Measuring ROI

Insights surfaces every input you need for a return-on-investment calculation, but it does not display a single "ROI" figure on screen. That final step is a short calculation you do yourself. The good news is that Outbound Virtual Agents offers one of the cleanest ROI stories in the whole Max platform, because the bookings are tagged and the revenue is directly attributable to the capability.

Note: The ROI method below is a recommended calculation built from the metrics ServiceTitan reports natively. Insights displays the inputs (Completed Job Revenue, Jobs Completed, response and conversion rates); it does not compute the ratio for you. Treat the formulas here as a measurement framework, not a native platform output.

Step 1 — Establish the return (attributed revenue)

Start from Completed Job Revenue for your chosen window, taken from the Jobs & Revenue section of the goal you are measuring. Use completed rather than booked revenue so you are measuring cash earned, not cash expected. This figure is the "revenue attributed" to Outbound Virtual Agents — it counts only jobs the AI agent originated, because those bookings are tagged.

Step 2 — Establish the investment (cost)

Total what the capability costs you over the same window. Depending on how your agreement is structured, this can include the subscription or per-usage cost of the virtual agent capability and any directly associated costs your CSM or PAM can confirm. Because pricing and packaging vary by agreement, confirm the exact figures with your CSM or PAM rather than estimating.

Step 3 — Calculate ROI

With return and investment for the same window, the standard calculation is:

ROI (%) = ((Attributed Revenue − Investment) ÷ Investment) × 100

A simpler companion figure many owners prefer is the revenue multiple:

Revenue Multiple = Attributed Revenue ÷ Investment

A multiple above 1.0 means the capability is returning more than it costs. Both figures use the same two inputs; the percentage frames it as a return, the multiple frames it as "dollars back per dollar spent."

Step 4 — Sanity-check the quality behind the number

An ROI figure is only as trustworthy as the attribution and completion behind it. Before you present it, confirm:

Attribution is clean. Jobs booked by the AI agent are tagged with the correct lead source. Missing tags understate your return.
You are using completed, not booked, revenue. Booked jobs that never complete are not return.
Escalation patterns are not quietly capping performance. Review escalation reasons — Escalated (Availability) points to Adaptive Capacity, Escalated (Skill/Handling) points to Voice Agent configuration in Contact Center, and Escalated (Negative Sentiment) or Escalated (Customer Requested) point to persona and rules tuning. A high availability-escalation rate means the agent is offering times your business does not actually have, which suppresses conversion and, with it, ROI.

A worked example

The figures below are illustrative, to show the method — they are not benchmarks.

Suppose over a 90-day window a Speed to Lead campaign shows Completed Job Revenue of $60,000, and your CSM confirms the capability cost you $12,000 over the same window.

ROI = (($60,000 − $12,000) ÷ $12,000) × 100 = 400%
Revenue Multiple = $60,000 ÷ $12,000 = 5.0×

You would then read that alongside the operational metrics — response rate, conversion rate, and the availability-escalation rate — to confirm the return is durable rather than a one-window spike.

Set your own benchmarks

ServiceTitan does not publish universal benchmarks for response rate, conversion rate, or ROI, because a healthy number depends on your trade, your lead sources, and your goals. Before a review, write down what success looks like for your business — for example, a target response rate or a minimum revenue multiple — and measure against those targets rather than an industry average. The most useful comparison is almost always your own trend over time: this window against the last.

Using what you see

Reporting only earns its keep when it changes a decision. A few patterns are worth acting on directly:

Compare response rates across campaigns to see which lead sources engage best. A campaign with a low response rate usually needs a message or cadence change, not more budget.
Read the funnel to locate the leak. Fix the biggest single drop first — and fix it in the system that owns it (Marketing for the top of the funnel, Adaptive Capacity or Contact Center for the middle).
Watch the Jobs Booked to Jobs Completed gap. A widening gap is booked revenue that is not converting to cash, and it deserves a look at scheduling and completion workflows.
Treat a rising Opt Out Rate as an early warning on message frequency or tone, before it pulls response rate down.
Bring the consolidated numbers to your CSM or PAM. They can compare your patterns against what they see across other ServiceTitan Max customers and help decide whether to expand campaigns to additional lead sources or goals. Any expansion has to be backed by Adaptive Capacity — Speed to Lead can only book into availability that actually exists.

The Insights page is your week-to-week view for this. Your CSM may also review broader ROI with you on your monthly Max call, putting your Outbound Virtual Agents numbers in the context of the rest of your business — so treat Insights as the cockpit you check between those conversations.

For the structured quarterly version of this review — including the sign-off steps and next-quarter decisions — see the ServiceTitan Max 90-Day ROI Review Template and Benchmark Guide, where attributed revenue compared against your investment is described as the cleanest ROI calculation available for Outbound Virtual Agents.

Documentation Index