AI appointment setting case study

75 appointments a week, booked automatically.

HostMetric built a 24/7 AI appointment-setting system for a high-volume energy-efficiency provider. It responds, qualifies, books, updates the CRM and gives the sales team the context needed to close.

Live system operating across SMS, CRM and calendar workflows.
Live lead journeyActive
EN
New energy-system enquiryReceived outside business hours
10:47pm
AI

Thanks for your enquiry. I can help organise a suitable appointment and collect a few details for the team.

InterestHeating + hot water
UrgencyHigh
AppointmentTuesday · 10:30am
CRM statusSynced
Booking completedCalendar and HubSpot updated
75Average appointments booked each weekLater operating period
23.48%Close rate at the increased volumeLater operating period
3xInstall rate for AI-booked leadsOriginal comparison cohort
40–60Previously missed leads recovered weeklyReported operational impact
The commercial problem

The leads were arriving. The coverage was not.

The business was generating approximately 100 inbound enquiries a day, but a meaningful share arrived overnight or on weekends. Demand existed; the operating model could not respond consistently while customer intent was still active.

01

Hot leads cooled before contact.

Enquiries waited for the next staffed period even when customers were actively comparing providers.

02

Representatives spent too much time chasing.

Highly trained sales staff repeatedly contacted people who would not respond or were not ready to proceed.

03

Bookings created downstream delays.

Appointment, payment, rebate and installation steps depended on manual administration and follow-up.

04

Customer context was fragmented.

Salespeople often needed to rediscover product interest, urgency and property information during the first call.

This was not a staff-performance problem. It was a process-coverage problem: immediate, consistent handling could not economically depend on someone always being available.

One connected conversion journey

The system carries the enquiry beyond the first reply.

Choose a stage to see what happens behind the conversation. Leave it running to follow the full journey automatically.

Lead intake

A new enquiry arrives.

A website form, campaign response or CRM event starts the workflow at any hour of the day.

Commercial purposeThe enquiry enters one controlled response path.
Workflow eventEnquiry
Live signal10:47pm · new heating and hot-water enquiry
ResponseImmediate
Next actionIn progress
Customer recordListening
Stage 1 of 7
What customers experienced

Real conversations showed more than instant response.

The live system handled qualification, changing availability and useful sales context. These anonymised reconstructions keep the tone and behaviour of the real threads without publishing private customer data.

Conversation outcome

The enquiry was handled before the team returned.

A late-night enquiry received an immediate response, selected a suitable appointment and entered the calendar without waiting for the next business day.

What the system captured
Product interestConfirmed appointmentCurrent system ageCross-sell opening
SMS conversationSynced
Assistant

Hi, saw your enquiry come through about a hot water system. Happy to line up a quick call so the team can look at what works and sort out the rebates. Would Tuesday morning suit?

Lead

Tuesday morning would be ideal.

Assistant

Tuesday at 10:30am it is — I've locked that in. Quick one so the team can come prepared: anything about the current system they should know?

Lead

It's ancient, and we're thinking about heating and cooling too.

Assistant

Good to know. I've flagged both so they can bring options to the call. You're all set for Tuesday at 10:30am — just sing out if anything changes.

More than a chatbot

The conversation was only one layer of the system.

Each responsibility was separated into a controlled module. That made the workflow easier to maintain, debug, measure and extend without allowing one failure to stop every other task.

01

Lead intake and response

Forms, campaigns, landing pages and CRM events enter one workflow, and the first reply adapts to source, product interest and time of day.

Context selected
02

Qualification

Product, property, urgency and readiness are converted into usable sales information.

Intent structured
03

Booking engine

Live availability, appointment rules, confirmations and rescheduling are handled together.

Calendar synced
04

Payment and reminders

Appointment reminders, payment prompts and rebate follow-up run on a schedule instead of depending on someone remembering.

Follow-up running
05

CRM orchestration

Contacts, notes, labels, outcomes and ownership are updated across the operating stack.

Record enriched
06

Governance

Opt-outs, uncertain cases, urgent requests and human escalation follow defined rules.

Boundary checked
System principleOne module handles the conversation. Another manages data. Another books. Another records the result.

The customer experiences one coherent assistant while the implementation remains modular and fail-safe behind the scenes.

Sales handover

The team did not just receive a booking. It received the context.

The CRM record tells the representative what the customer wants, why they enquired, how urgent the situation is and what should happen next.

Less time reconstructing the conversation
Clearer prioritisation for urgent enquiries
Cross-sell and product-interest signals
Appointment, payment and reminder status
Immediate opt-out and compliance handling
Contact recordQualified appointment
AI-booked
EN
Energy-system enquiryConversation and booking synced
Product interestHeating + hot water
AppointmentTuesday · 10:30am
Current situationAgeing system
Property contextSplit-level home
UrgencyHigh priority
Cross-sellHeating and cooling
Integration note

Customer confirmed the appointment by SMS. Product interest, property context and urgency recorded. Sales representative to review heating, cooling and hot-water options before the call.

Human boundaries

Automated where repeatable. Escalated where judgement matters.

The assistant does not need to solve every enquiry. It needs to recognise what it can complete, what it should clarify and what a person must handle.

Automate

High-volume tasks with known rules

  • Immediate acknowledgement
  • Approved qualification questions
  • Calendar availability and booking
  • Appointment and payment reminders
  • CRM notes, labels and outcomes
  • SMS opt-out processing
Escalate

Cases requiring judgement or authority

  • Uncertain technical questions
  • Urgent or safety-related situations
  • Non-standard commercial requests
  • Complaints and sensitive cases
  • Requests outside approved boundaries
  • Low-confidence intent or data
Original comparison cohort

AI-booked leads converted materially better.

The original analysis compared every lead in the client's account against the leads that completed an appointment through the assistant, measured across the same window in their HubSpot portal.

Deal creation rate

+105%
All leads33.08%
AI-booked67.81%

Deal close rate

2.8x
All leads8.66%
AI-booked24.76%

Install rate

3x
All leads5.31%
AI-booked16%
AI-booked leads achieved a 2.8x close rate and a 3x installation rate in the original measured cohort.

These were largely after-hours enquiries that were at elevated risk of delayed or inconsistent follow-up.

Later operating period

The first deployment was not the finished system.

HostMetric continued refining routing, prompts, qualification rules and operational logic after launch. Booking volume increased while the close rate remained above 23%.

≈50Weekly appointments at the earlier winter peak
≈75Average weekly appointments in the later period
23.48%Close rate at the higher booking volume
Weekly appointmentsContinued optimisation
+≈50%
≈50≈75Earlier peakLater period
Illustrative · two measured points, six months apart
Associated weekly closed deal value≈$158,000

75 weekly appointments × 23.48% close rate × the client's $9,000 average deal value. Associated closed value, not a claim of incremental revenue.

Commercial fit

This works when lead handling is a process—not just a message.

The strongest opportunities combine meaningful lead volume, higher customer value, qualification complexity and an operating journey that currently depends on manual follow-up.

01

Meaningful inbound lead volume

The commercial case becomes stronger when delayed follow-up affects dozens or hundreds of enquiries each month.

02

Leads arrive outside staffed hours

Evenings, weekends and campaign spikes create response gaps that are expensive to cover manually.

03

Qualification matters before booking

The team needs product, property, urgency or readiness information—not simply another calendar entry.

04

Several systems must agree

Forms, CRM records, calendars, SMS, payment and staff handover need to behave like one process.

05

Sales time is being consumed by chasing

Representatives spend too much time contacting unresponsive leads or reconstructing customer context.

06

Bookings leak before the final outcome

Delays, missing reminders and inconsistent follow-up affect attendance, payment or installation progression.

A useful starting questionHow many enquiries each month wait, go unqualified or disappear before a confirmed sales conversation?
Lead-response review

See where your lead journey is slowing down.

HostMetric will review how enquiries currently move from submission to contact, qualification, booking and sales handover.

Where response and follow-up currently break
What can be automated safely
What should remain human
Which systems and data must be connected
Where the likely commercial opportunity sits
No automatic system access.Any data review, connection or workflow change is discussed and approved separately.
Request a review

Tell us where leads currently stall.

By submitting this form, you agree that HostMetric may contact you about this enquiry. No customer communication or system change is launched without a separate scope and approval. See our privacy policy.