agentimus.io Book a discovery call

Case studiesHome services, Newark

AI receptionistQuote generatorInvoice automationCRMData minimisation

How a Newark plumbing company cut quote-to-payment from 8 days to same-day

An 8-truck mixed residential and commercial operation in Newark was quoting by hand, invoicing days late, and tracking customers in a shared Google Sheet. We rebuilt their whole revenue back office with automation. Here's the result.

+41%Quote close rate
1.2 daysAverage quote to payment (was 8)
22 hrsOffice admin saved per week
$28,600Additional closed revenue in 60 days

Confidential by request. Company name and staff names withheld under NDA. Location: Newark, NJ. Installed Q4 2024.

Growing revenue. Growing chaos. No system holding it together.

This company had grown from 3 trucks to 8 in two years. The back office hadn't grown with it. Everything after a job was still manual, and getting slower as volume increased.

  • Quotes took 1 to 3 days to send

    The owner wrote quotes by hand in a Word doc, estimated materials from memory, and emailed them whenever he had time. By the time the quote arrived, half the customers had moved on.

  • Invoices sent 5 to 8 days after the job

    There was no invoicing workflow. Jobs got invoiced whenever the office caught up. One month they had $34,000 in completed work that hadn't been billed.

  • Customer data in a shared spreadsheet

    Job history, contacts and follow-up notes all lived in one Google Sheet that three people updated inconsistently. No pipeline, no follow-ups, no way to see which leads were going cold.

The math: their average job value was $820. With a 28% quote close rate and a 5 to 8 day payment lag, they were losing quotes to speed every week and sitting on cash already earned. Fixing close rate and invoice speed alone was worth an estimated $35,000+ in recovered monthly revenue.

Four automations. Installed in 14 days.

We connected call intake, quoting, invoicing and the customer pipeline into one automated system. Each piece feeds the next.

Automation 1 of 4

AI voice receptionist: every call captured, qualified and logged

We installed a 24/7 AI voice agent on the main business line. It answers every call in under 5 seconds, weekends included. For new leads it collects job type, property address, residential or commercial, and urgency, and that goes straight into the CRM as a new lead. For commercial service contracts, it captures company, scope and contact details and flags the record for the owner to call back within the hour.

748Calls handled automatically (60 days)
94New leads logged to the CRM
100%Calls with job details captured
11Commercial contracts flagged
Inbound call: commercial service enquiry+1 (973) 555-0381, Dec 5, 2024, 10:22 AM, 2m 18sCRM lead created
AI, 10:22:09

Thank you for calling. I'm the scheduling assistant. Are you calling about a residential job, a commercial property, or a service contract?

10:22:27

Commercial. We manage a 12-unit apartment complex on Branch Brook Ave. Looking for a plumber on a service contract basis: monthly maintenance and on-call.

AI, 10:22:41

Perfect, we do service contracts for multi-unit residential. Can I get your name, the property address, and the best number to reach you? The owner will call you back within the hour with details and pricing.

10:22:58

Marcus ████████, property manager. 447 Branch Brook Ave, Newark. You can reach me at this number.

CRM lead created: Marcus ████, commercial, 447 Branch Brook Ave. Priority high. Owner alerted 10:23 AM.

Lead name partly redacted. Every qualifying call creates a CRM record automatically, and the owner sees the lead before they hang up.

Automation 2 of 4

Automated quote generator: sent in minutes, not days

After a site visit or a qualifying call, the tech or office fills in a simple job form: scope, materials, labour estimate, any extras. The system generates a branded quote PDF, calculates the total with markup, and sends it by WhatsApp and email, usually within 15 minutes of the visit ending. The customer approves with one tap, which creates the job in the schedule. If a quote sits unread for 24 hours, a follow-up goes out automatically.

14 minAverage quote sent after a site visit
69%Quote close rate (was 28%)
91%Quotes approved the day they're sent
$1,320Average quote value (60 days)
Service estimate, quote #Q-2024-0847Dec 5, 2024. Valid for 7 daysApproved 11:04 AM
Prepared for ████ ██████████312 Grafton Ave, Newark, NJ. Water line replacement
DescriptionQtyAmount
Labour: water line excavation and replacement (est. 6 hrs)1$840.00
3/4" copper supply line (40 ft) and fittings1$310.00
Shut-off valve replacement (main)2$180.00
Permit and inspection fee1$120.00
Total estimate$1,450.00

Customer tapped Approve on the WhatsApp link.

An auto-generated quote. Customer name redacted. Sent 14 minutes after the tech left the site and approved 42 minutes later on WhatsApp.

Automation 3 of 4

Invoice and payment automation: billed the day the job closes

The moment a tech marks a job complete, the system generates and sends the invoice, pulling from the approved quote, any change orders, and the tax rate for the property's zip code. It arrives by WhatsApp and email with a payment link (card, ACH or Zelle). No payment in 48 hours? A polite reminder. A second one at 5 days. The office never touches it unless a customer escalates.

1.2 daysAverage days to payment (was 8)
$0Unbilled completed work (was $34k)
94%Invoices paid within 48 hours
0 minOffice time per invoice
Invoice dashboardDecember 2024
CustomerAmountSentStatus
████ ██████Water line replacement, 312 Grafton Ave$1,450Dec 5, 3:12 PMPaid
██████████ Property MgmtMulti-unit drain clearing, Branch Brook Ave$2,800Dec 5, 4:47 PMPaid
████ ████████Boiler pipe repair, 88 Clinton Ave$680Dec 6, 9:30 AMPaid
████████ Associates LLCCommercial bathroom refit, McCarter Hwy$4,200Dec 6, 11:15 AMPending
All customer names redacted. Every invoice shown was generated automatically within minutes of the job closing. No one in the office typed a single one.

Automation 4 of 4

CRM pipeline: every lead tracked, every job stored

Every call, quote, job and payment now flows into one CRM pipeline. The owner sees at a glance how many new leads came in this week, which quotes are pending, which jobs are in progress and which invoices are outstanding. Past customer history is one click away, so when a client calls about a new job, the tech already knows the property and the last service. Quotes not approved in 5 days get a follow-up automatically.

94Leads logged automatically in 60 days
0Leads lost to no follow-up
22 hrsAdmin time saved per week
$28.6kAdditional closed revenue (60 days)
CRM pipelineWeek of Dec 2 to 8, 2024
New leads (7)
Commercial enquiryBranch Brook Ave, 12-unit
Residential leakIrvington, after-hours call
Drain backupEast Orange, WhatsApp
Quote sent (4)
Water line replacement$1,450, sent 14 min ago
Boiler refit$3,200, follow-up sent
Sewer line scope$890, awaiting approval
In progress (5)
Commercial bathroomMcCarter Hwy, day 2 of 3
Drain clearing x4 unitsBranch Brook, today
Boiler pipe repairClinton Ave, in progress
Invoiced and paid (18)
Water line, $1,450Paid same day
Drain clear, $2,800Paid in 4 hrs
Boiler, $680Paid next day
One week of the pipeline. The Google Sheet is gone. Every lead, quote, job and payment is in one place, updated automatically as the work moves through.

60 days after go-live

Measured from the first full week of operation.

+41%Quote close rate increase (28% to 69%)
1.2dAverage quote to payment, down from 8 days
22hOffice admin hours saved every week
$0Unbilled completed work (was a $34,000 backlog)
$28,600Additional closed revenue in 60 days from better close rate and speed
94%Invoices paid within 48 hours, no chasing
14 minAverage time from site visit to quote sent (was 1 to 3 days)

Before and after

MetricBeforeAfter (60 days)
Time from visit to quote sent1 to 3 daysAbout 14 minutes
Quote close rateAbout 28%About 69%
Days from job complete to invoice sent5 to 8 daysUnder 5 minutes
Unbilled completed work at any time$20k to $34kAbout $0
Office time on quotes, invoices and follow-upsAbout 4 hrs a dayUnder 20 min a day

How the data is handled

No health data here, but plenty of personal data: names, home addresses, phone numbers, access notes and payment details. Before this build, all of it sat in a shared spreadsheet that anyone in the business could open.

Where data lives and is processed
Customer and job records in the CRM. Invoices and payments in the accounting and payment tools. The shared spreadsheet was retired.
What the AI never handles
Card numbers. Customers pay through hosted payment links, so card data stays with the payment provider. The quote generator works from the job details and the price list, not the customer's payment history.
Who can access it
Office staff have roles in the CRM. Technicians see the jobs assigned to them, not the whole customer list.
How long it's kept
Call recordings and transcripts are kept for the period the owner set. Customer and invoice records follow the company's own retention for tax and warranty purposes.
What's logged
Every call, quote, invoice and payment status change, with who or what made it.

Rules it maps to

  • PCI DSSHosted payment pages keep card data out of the company's own systems, which narrows the scope of PCI DSS. PCI DSS
  • New Jersey Data Privacy ActData minimisation and purpose limits. Whether the Act applies depends on volume thresholds. The design follows its principles either way. NJDPA
  • Call disclosureCallers are told the call is handled by an AI assistant and recorded.

This section describes how data moves through the system as delivered. Mapping a control to a rule shows which requirement it supports. It is not a statement that the client is compliant: that judgement belongs to the client and its advisers.

"I had $34,000 in work I'd already done that hadn't been invoiced yet. That was just sitting there. The first thing this system did was clear that backlog, automatically. I didn't have to do anything. Now we invoice before the truck is back in the lot. I don't know how we functioned before."

Business ownerPlumbing company, Newark, NJ. Name withheld by request.

See all case studies

Services behind this system

More case studies

All ten case studies

Home services, South Jersey

How a South Jersey plumbing company went from missing 60% of after-hours emergencies to none, with an AI phone agent that answers in under 5 seconds and passes the on-call tech only what they need.

Read the case study

Car rental, Dubai

How a Dubai car rental fleet moved customer chats off staff phones, ended double bookings and added 38 confirmed bookings a month, with ID documents kept out of chat history.

Read the case study

Car rental, Los Angeles

How a Los Angeles luxury car rental lifted booking conversion 22% and avoided $8.4k in disputes, with licence checks, timestamped inspections and customer documents in one controlled place.

Read the case study

Tell us what you want AI to do, and the rules it must follow.

On a free 15-minute discovery call we ask what you want to build or already run, what data it involves and which regulations apply. You leave knowing what to build first and where the risks are. No obligation.

  • Pick a time that suits you
  • 15 minutes with an engineer
  • Clear next steps
Book a free discovery call Email us

Prefer to write first? hello@agentimus.io, or WhatsApp. We're glad to sign an NDA before you share details.

Or send us a message

Only your email is required. We use these details to reply to you. Privacy policy

Book a discovery call