Lesson 7 of 7 · 6 min

Closing the 30-day pilot

Close a 30-day pilot with clear terms and your own price.

Read the lesson

Goal: Close a 30-day pilot with clear measures, a clear price and a clear day 30 decision, all in writing.

You need: A finished reverse demo. The owner's discovery numbers. Package, Price and Deliver lesson 4 for the install.

Why this matters

A pilot lowers the owner's risk. They do not commit to a year. They commit to 30 days and two numbers. If the numbers move, the next step is easy. If they do not, both of you walk away with the truth.

A pilot also protects you. Without written measures, day 30 turns into a debate about feelings. "It seemed fine" does not sell a retainer. "Answer rate went from 86% to 97%" does.

Most pilots fail for one of three reasons. The scope was too big. The measures were never agreed. Or nobody decided what happens on day 30. This lesson closes all three gaps before day one.

What goes in the pilot

Scope: Catch only. The pilot runs the Never Miss a Job fix. That is the AI voice agent on missed calls, plus missed-call text-back. Do not add reminders, quotes or reviews. A small scope goes live fast and gives a clean result.

Two measures, agreed before day one.

  1. Answer rate. Calls answered divided by total calls. Find it in Analytics → Calls → Success Rate.
  2. Jobs booked by the AI each week. Count the jobs the agent booked, week by week.

Do not measure reviews, quotes or revenue in the pilot. They take longer than 30 days to move. Watch them later.

The baseline. Write the starting numbers from the audit. For example, a 14% miss rate means an 86% answer rate today.

The target. Set it with the owner, in numbers. Pick something the fix can reach in 30 days.

The price. You choose the pilot price. Some Operators run it free. Some charge a set fee. Some charge the setup fee and credit part of it to the retainer. ServiceAgent does not set or collect your fees. Whatever you choose, say who pays the ServiceAgent plan and credits during the pilot.

The account model. An agency-run pilot lives in an organization you create and pay for. A client-signed pilot means the owner signs up and pays ServiceAgent. Agency-run is simpler for a short pilot, because the owner signs nothing with ServiceAgent.

The timeline. The install takes about 7 days. The 30 days count from go-live, not from signing.

Fig 1

From yes to the day 30 decision

  1. SigningOne-page pilot agreement
  2. Days 1 to 7Install; forward only after the day 6 tests
  3. Go-liveDay 1 of the pilot; note the date
  4. Each weekAnswer rate and AI-booked jobs in a three-line update
  5. Day 30The decision written in the agreement
The 30 days start at go-live, not at signing.

The day 30 decision. Write it down now. If the targets are met, you present a retainer proposal. If not, you share what you learned and part on good terms.

Fig 2

The one-page pilot agreement

  • Scope: Catch only
  • Measures: answer rate and AI-booked jobs a week
  • Baseline from the audit
  • Target in numbers
  • Price, and who pays the plan and credits
  • Account model: agency-run or client-signed
  • Timeline: 30 days from go-live
  • The day 30 decision
Every line is agreed before day one.

By size. For an owner-operator, run the pilot on all missed calls. For a fleet, start with after-hours calls only, so the office team keeps its routine. For a multi-location group, pilot one location. Then roll out to the others if it works.

Fig 3

Pilot scope by business size

Owner-operator
  • All missed calls
Fleet
  • After-hours calls only
  • The office keeps its routine
Multi-location
  • One location
  • Roll out if it works
Keep the scope small, so it goes live fast and gives a clean result.

How to ask

Ask at the end of the reverse demo, while the booked test job is still on the screen. Keep it plain: "Want to try this for 30 days on your missed calls?" Then stop talking. If they hesitate, ask what worries them and use lesson 6. Do not fill the silence with extra features. A clear, small offer is easier to say yes to than a big one.

Build it in ServiceAgent

  1. Open the Organisations page (/multi-account) and choose Add organisation for an agency-run pilot.
  2. Choose the plan. Core fits most pilots. It is $39 a month billed yearly or $49 month to month.
  3. Turn on Auto-reload Credits in Billing → Payment settings. Set Reload when below and Add amount, so the agent never runs dry.
  4. Follow the 7-day install in Package, Price and Deliver lesson 4. Do not forward the line until the day 6 tests pass.
  5. On go-live day, note the date. Open Analytics → Calls and record Success Rate for the week before, if call data exists.
  6. Each week, record answer rate and AI-booked jobs in a note on the deal. Send the owner a three-line update.
  7. Check it works. On day 7 of the pilot, compare the week's answer rate with the baseline. If it has not moved, check the forwarding first.

Worked example: Northside Plumbing's pilot

Carol agrees to a pilot after the demo. You write a one-page pilot agreement together.

Baseline. About 25 calls a business day and a 14% miss rate. Answer rate today: 100% − 14% = 86%.

What the fix should reach. The AI answers calls the team misses, up to 10 at a time. So the answer rate should rise close to 100%. You agree a target of 95% or higher.

AI-booked jobs. About 77 calls a month go unanswered today. That is about 77 ÷ 4.4 = 17.5 a week, using 4.4 weeks a month. At a 42% booking rate, that is about 7 jobs a week. You agree a cautious target of 4 or more AI-booked jobs a week.

Fig 4Example numbers

Northside's pilot targets

Answer rate today, 100% − 14%86%
Missed calls a week, 77 ÷ 4.417.5
×Booking rate42%
=Possible AI-booked jobs a weekabout 7
Targets agreed95% answered; 4+ jobs a week
Set targets the fix can reach in 30 days.

Price and model. You choose a $500 pilot fee, with the build included. The account is agency-run on Core, billed month to month. You pay the $49 plan and any credits during the pilot.

Day 30. If both targets are met, you present Never Miss a Job at $397 a month. If not, you share the data and stop.

Fig 5

The day 30 decision

  1. Day 30: read the two measures
  2. Were both targets met?YesPresent the retainer proposalNoShare the data and part on good terms
  3. Northside: 97% answered, 5 AI-booked jobs a weekBoth targets met
Write the decision into the agreement now, not on day 30.

After 30 days, the numbers come in at 97% answered and an average of 5 AI-booked jobs a week. Both targets are met. The retainer proposal opens with those two numbers.

Fig 6Example numbers

Northside's answer rate

Before the pilot86%
Target95%
After 30 days97%
The retainer proposal opens with these numbers.

Common mistakes

  • Piloting everything. The result gets muddy and slow. Fix: Catch only.
  • No written measures. Day 30 becomes a debate. Fix: two measures, a baseline and a target, signed before day one.
  • Starting the clock at signing. The install eats a week of the pilot. Fix: count 30 days from go-live.
  • Forgetting day 30. The pilot drifts into free work. Fix: write the day 30 decision in the agreement.
  • Measuring revenue. Thirty days is too short for it to be fair. Fix: measure answer rate and AI-booked jobs.

Do this now

Write a one-page pilot agreement for one real or practice prospect. Include scope, two measures, baseline, target, price, account model, timeline and the day 30 decision. Time: 20 minutes.

Practice the conversation

  1. Operator Carol, let's try it for 30 days on missed calls only.
  2. Owner What do I have to do?
  3. Operator One kickoff call and a test session. I do the build.
  4. Owner And how do we know if it worked?
  5. Operator Two numbers. Your answer rate is 86% today. We aim for 95% or more.
  6. Owner And the second?
  7. Operator Jobs the AI books each week. We aim for at least 4.
  8. Owner What if it doesn't hit them?
  9. Operator Then I show you the data and we stop. No hard feelings.
  10. Owner And if it does?
  11. Operator I bring you a monthly plan on day 30. You decide then.

Quick check

  1. A plumbing owner says yes to a 30-day pilot of Never Miss a Job. Which measures do you agree in writing before day one?
    • Credits used and AI call minutes each week
    • Answer rate and jobs booked by the AI each week
    • Whether the owner feels it went well on day 30
    • New reviews and accepted estimates
  2. You sign a pilot on the 1st. The install finishes and the line goes live on the 8th. When is day 30?
    • The 30th
    • The 31st
    • 30 days after the 8th
    • 30 days after the first booked job
  3. An owner misses 20% of 15 calls a business day. Using 22 days and 4.4 weeks a month, about how many missed calls happen each week?
    • 10
    • 15
    • 20
    • 66

ServiceAgent UniversityLearn to run a service business on autopilot. Free courses, Blueprints and certification for Service Operators.

Learn

Get certified

Community