Your business is losing revenue to missed calls, cold leads, and manual follow-ups.
I build complete automation systems that recover that revenue — from first contact to repeat client. No-code, nothing new for your staff to learn. Just a system that runs.
Built withMake.comn8nGoHighLevelZapierClaudeGeminiOpenAI
Revenue on line 1
9:41
Incoming call
Revenue calling…
a lead who buys from whoever answers
Call answered.
Audit request received — talk soon —
That's the leak. 78% of customers hire whoever answers first — they're already dialing your competitor.
Missed calls are where revenue dies. Don't miss this one — tap the green button.
From a former accountant
The leak ledger.
I spent five years as a founding accountant — I know exactly where service businesses bleed revenue, because I was the one tracking the numbers. The leaks are the same in every trade; they never show up as a line item, so nobody fixes them. Here's what that statement would look like if they did.
Revenue leak statementany service business · any given month
Calls missed while the crew is on the job(78% hire first responder)
Quotes sent once, then forgotten(3× left unclaimed)
No-shows nobody rescues($500–$700 each)
Rebooking windows that quietly pass(60% of windows missed)
Balance: revenue you already earned( leaking )
What I do: build the system that recovers every line above.
Whole-workflow systems
Not a chatbot. Not one zap. The whole workflow.
Three complete systems — each covers first contact → booked → follow-up → review → reactivation. Scoped to your business, handed off running.
Roofing · HVAC · Plumbing · Electrical
Contractor Lead Rescue & Quote Follow-Up
Your crew's on a roof when a $15,000 job calls. Nobody answers — so they call your competitor next. Meanwhile quotes sit for days and old leads die untouched.
The mathOne recovered $8K–$25K job per month repays the system in full.
24/7 AI receptionist + missed-call text-back — every call, form & LSA lead in 60s
After-hours emergency capture & routing
Automated quote follow-up — day 1, 3, 7
Job-complete review request + referral ask
Dead-quote reactivation to dormant leads
Med spas · Aesthetics
Med Spa Client Lifecycle & Reactivation
A client gets Botox, means to rebook in 3 months, and forgets — 60% of rebooking windows end this way. No-shows run 15–30% at $500–$700 each.
Deposit-confirmed reminders — no-shows drop ~25% → 8–12%
The mathOne reactivation campaign on 400 lapsed clients: $20K–$36K, zero ad spend.
Dental clinics
Dental Patient Recovery
No-shows cost the average practice an estimated $105K–$240K a year; missed calls another $64K–$266K. Meanwhile staff dial reminders by hand and lapsed patients hear nothing.
Missed-call text-back with booking link — Viber, WhatsApp, SMS
Invoice Follow-UpPolitely chases unpaid invoices, stops the instant payment lands.stops itself on payment
Post-Job Review RequestsFires the review ask right after job completion, nudges the quiet ones.reviews on autopilot
AI Lead Triage & Auto-ResponseReads free-text inquiries and routes emergencies first.triage without a human reader
After-Hours Emergency CaptureKnows 3 AM from 3 PM — wakes the on-call only for emergencies.emergency jobs run 2–3× value
Post-Job Review RequestTurns completed jobs into Google reviews automatically.72% review if asked; 14% are
Med Spas4 builds
Instant Lead QualifierScores every lead HOT / WARM / COLD with the next step.HOT → callback in 15 min
No-Show Recovery & RebookingStarts a rebooking sequence the moment an appointment is missed.$500–$700 saved per rescue
Treatment-Clock RebookingPer-client timers nudge the rebooking exactly when results fade.>50% of windows normally missed
Weekly Performance DigestMonday-8am digest — bookings, revenue, no-show rate — with a data-feed alarm.catches a bad week on Monday
Dental Clinics2 builds
24/7 AI Front-Desk (RAG)Answers from the clinic's own prices & policies — can't invent a wrong one.answers at 8pm & on Sundays
Dental Booking Front-DoorOne webhook, four routes, with a duplicate check that can't double-book.duplicate-proof, verified live
Insurance Agencies3 builds
24/7 Insurance ConciergePlain-English quotes from controlled pricing logic — not hallucinated.replies in under 3 seconds
Quote → CRM + Instant Follow-UpQuote requests hit the CRM and get a reply in under a minute.40% go cold in 15 min
Policy Renewal SequenceDisciplined 60 / 30 / 7-day renewal cadence across the whole book.attacks 10–15% annual attrition
Growth & Ops2 builds
Local Prospect FinderScrapes and scores local businesses into a live prospect list.list-building: hours → minutes
Global Error HandlerCatches any workflow failure and emails the details instantly.no silent failures
About
Numbers first. Automation second.
I spent 5 years as the founding accountant for two companies — financial systems from scratch, BIR compliance, international transfers in four currencies. I know what broken operations cost a business, because I tracked the numbers.
Now I build automation systems that fix those processes — whole-workflow builds for service businesses, first contact to repeat client, on Make.com, n8n, and GoHighLevel.
I'm not a tool trainer. I design, build, and deploy. When I hand off a system, it runs — 24/7, without your staff babysitting it.
The file on me
NameErjan Paul Poblete
BaseQuezon City, Philippines
HoursI work your business hours — clients in the US, Canada, UK & Australia
Texts every missed caller back within seconds — before they dial the next roofer on Google.
The workflow itself — GoHighLevel workflow canvas
The problem
Roofing crews are on roofs, not by the phone. An unanswered caller — a homeowner ready to book an estimate — simply dials the next roofer. The first contractor to respond wins the job, and most missed calls are never followed up at all.
The system
Fires on any inbound call that doesn't connect — no-answer, busy, voicemail, or abandoned mid-ring
Duplicate guard: a caller who's already been texted is never messaged twice, even if they ring repeatedly
A 12-second pause before sending, so the text reads like a person reacting — not a robot
Instant SMS with a self-serve booking link to schedule an estimate
Staff alert on every miss, so a team member can also follow up personally
What changes
Recovers an estimated 25–45% of missed calls that would otherwise go to a competitor
Responds in seconds, 24/7 — nights and weekends included
A missed call becomes a booked estimate with zero effort from the crew
Why it matters78% of customers hire whoever responds first.
Documented portfolio build · full write-up with workflow screenshots on request
Build 002 · GoHighLevel · Roofing / Contractors
Lead Capture → Pipeline → Follow-Up
An end-to-end lead engine that captures, organizes, and follows up — and knows when to stop.
The workflow itself — GoHighLevel workflow canvas
The problem
Roughly 40% of quote requests go cold within 15 minutes when no one replies — and a crew on a job site isn't watching the website form. The lead was paid for, then lost in silence.
The system
Website estimate form creates the lead and drops a deal into the pipeline at "New Lead" — nothing is ever lost
Personalized text + branded email within seconds, with a link to book a time
3-touch follow-up: nudges on day 1 and day 3 if they don't book
Smart stop: before every touch it checks whether a rep has picked up the deal — the moment a human engages, the automation goes quiet
What changes
Every quote request answered in seconds — the speed-to-lead race is won by default
Every lead visible in a pipeline; cold leads recovered with zero manual follow-up
Professional, never spammy — a worked lead is never pestered by a robot
Why it matters40% of quote requests die in the first 15 minutes of silence.
Documented portfolio build · full write-up with workflow screenshots on request
Build 003 · GoHighLevel · Roofing / Contractors
Appointment Reminder Ladder
Online booking plus a reminder ladder that keeps every estimate visit on track.
The workflow itself — GoHighLevel workflow canvas
The problem
An estimate no-show means a crew blocked time and drove out for nothing — and chasing confirmations by phone eats the office's whole day. Most no-shows aren't deliberate; people just forget an appointment they booked days earlier.
The system
Self-service booking on a branded calendar — the slot reserves instantly, no phone tag
Instant text + email confirmation with a one-tap reschedule option
Automatic nudges at 24 hours, 2 hours, and 30 minutes — the last doubles as "estimator on the way"
Reschedule-aware: reminders are tied to the live appointment time, so if the homeowner moves the slot, the reminders move with it
What changes
Fewer no-shows and wasted crew trips
Zero manual confirmation calls — the calendar runs itself
A small contractor delivers the on-time experience of a big operation
The ladder24h · 2h · 30min — and it follows every reschedule automatically.
Documented portfolio build · full write-up with workflow screenshots on request
Build 004 · GoHighLevel · Roofing / Contractors
Invoice Follow-Up
Chases politely, stops the instant it's paid, and knows when to hand off to a human.
The workflow itself — GoHighLevel workflow canvas
The problem
Chasing unpaid invoices is the job every contractor dreads — awkward, inconsistent, and it eats the owner's evenings. Some customers get chased, some slip through, and the ones who already paid sometimes get a reminder anyway.
The system
Starts when the invoice goes out — a short grace period, then a friendly text + email reminder
A second, firmer nudge a few days later if it's still open
Stops the moment they pay — payment instantly pulls the customer out of the sequence
The deal moves itself to "Paid" in the pipeline, so the books always reflect reality
Still unpaid after the nudges → a task for the owner to make a personal call
What changes
Invoices get paid faster with zero awkward manual chasing
A paying customer is never nagged
The owner's time goes only to the few accounts that genuinely need it
The mechanicPayment kills the sequence instantly — automation chases, humans close.
Documented portfolio build · full write-up with workflow screenshots on request
Build 005 · GoHighLevel · Roofing / Contractors
Post-Job Review Requests
A review engine that fires at the perfect moment and nudges only the people who went quiet.
The workflow itself — GoHighLevel workflow canvas
The problem
Google reviews are the biggest driver of local ranking and trust — but the ask almost never happens. The crew finishes and moves to the next job, and the perfect window (the few hours right after completion, while the customer is admiring the new roof) closes.
The system
Triggers the moment the office marks a job complete — no one has to remember
A short wait, then a friendly text + email with a one-tap Google review link
Watches engagement for up to three days
One polite follow-up — sent only to customers who ignored the first ask
What changes
A steady, automatic flow of fresh Google reviews
Every ask lands at the ideal moment, with zero effort from crew or office
No customer is ever over-messaged
Why it mattersReviews are the engine of the local map pack — this makes the ask automatic.
Documented portfolio build · full write-up with workflow screenshots on request
Build 006 · Make.com · Dental clinics
Dental Booking Front-Door
One webhook, four routes — a production-grade router, not a happy-path dispatcher.
The workflow itself — Make.com scenario canvas
The problem
Every time a patient books, reschedules, or cancels online, someone at the front desk has to notice, copy the details across, email the patient, and tell the dentist. If the desk is busy, it waits — sometimes it's forgotten entirely.
The system
Catches every booking event the instant it happens and routes it — new, rescheduled, or cancelled
Checks for duplicates first: booking platforms redeliver events, so a repeat is recognized and silently ignored instead of double-booked
Updates the record, emails the patient, and notifies the front desk — all three, every time, in seconds
A genuine fallback path: an unexpected event type is logged and flagged for a human, never silently dropped
Failed runs land in a replay queue — a hiccup becomes a one-click retry, not a lost booking
Verified
Every path tested live — including deliberately firing a duplicate event to confirm the safeguard holds under the exact condition it exists to catch.
The edgeIdempotency checked, not assumed — zero double-bookings by design.
Documented portfolio build · full write-up with canvas screenshots on request
Build 007 · Make.com · Med spas
Weekly Performance Digest
Monday-morning numbers with zero manual pulling — and an alarm when the data feed dies.
The workflow itself — Make.com scenario canvas
The problem
Most small med spas never see how last week actually went, because getting the numbers means counting by hand. A rising no-show rate or a service that quietly stopped selling only surfaces at month-end — when the lost revenue is already lost.
The system
Wakes on schedule every Monday at 8am and reads the last 7 days of bookings on its own
Computes the metrics in-flight — total bookings, completed revenue, no-show count and rate — across every appointment in one pass, no spreadsheet round-trip
Delivers three ways: full email digest, a headline ping to the owner's phone, and a durable logged record that builds week-over-week history
Never sends silence: an empty result fires a distinct "check that the booking system is syncing" alert — a broken feed can't masquerade as a slow week
Verified
Illustrative week from test data: 21 bookings, $8,285 completed revenue, 3 no-shows, 14.3% no-show rate — computed and delivered untouched.
Why it mattersNo-show rates drop from 15–30% to 8–12% when someone's actually watching.
Documented portfolio build · full write-up with canvas screenshots on request
Build 008 · n8n (self-hosted) · Dental clinics
24/7 AI Front-Desk (RAG)
A grounded, three-mode AI assistant — it answers from the clinic's own documents or not at all.
The workflow itself — n8n workflow canvas
The problem
A dental front desk answers the same questions hundreds of times a month — at 8pm, on Sundays, mid-procedure. The usual fix, a generic AI chatbot, creates a worse problem: it makes up prices with total confidence. That destroys trust.
The system
Inform — retrieval-augmented (RAG): searches the clinic's actual documents (services, prices, policies, aftercare) before answering; if it's not in the docs, it says so and offers the team — it cannot answer from internet memory
Book — captures name, contact, service, and preferred time, then logs a booking request for the front desk to confirm
Escalate — clinical or urgent messages (swelling, severe pain) stop the conversation and alert staff immediately; it never attempts medical advice
Two workflows: document ingestion (chunk → embed → vector store) and the assistant agent that runs on every message
What changes
Accurate answers 24/7 on any channel that can send a webhook — website chat, Messenger, WhatsApp
Booking leads captured complete, with every answer traceable to its source
Update a document, re-run ingestion — the assistant knows the new information in minutes
The edgeIt can't invent a price — every answer is retrieved, not remembered.
Live showcase build, verified with real test conversations · full write-up on request
Build 009 · n8n (self-hosted) · Med spas
Instant Lead Qualifier
One scoring brain shared by every intake channel — HOT / WARM / COLD with the next step attached.
The workflow itself — n8n workflow canvas
The problem
Med spa leads arrive from everywhere — website, Instagram DMs, Facebook, walk-ins — and they're not equal. When every channel scores leads separately, the logic drifts, hot leads sit in the same queue as cold ones, and the front desk triages by hand.
The system
A single lead-scoring engine that every channel calls — the rules live in exactly one place
Validates each entry first; a lead missing treatment interest or contact info gets flagged, not junk-scored
Scores treatment value, booking readiness, and membership intent → HOT, WARM, or COLD
Returns the verdict with the recommended action: HOT → call within 15 minutes + provider alert; WARM → nurture sequence; COLD → newsletter
What changes
The highest-value inquiries get human speed while the person is still interested
Change the rules once — every intake updates instantly, no drift
A new channel is a pointer to the engine, not a rebuild
The ruleHOT lead → callback inside 15 minutes, every channel, every time.
Documented portfolio build · full write-up on request
Build 010 · n8n (self-hosted) · Home services
AI Lead Triage & Auto-Response
Reads free-text service requests and tells a burst pipe from a dripping faucet — instantly.
The workflow itself — n8n workflow canvas
The problem
Home-services leads arrive as free-text messages, not clean form fields. An emergency and a routine job look identical until someone reads them — and slow triage loses emergency jobs to the first competitor who responds.
The system
Lead arrives via webhook from any form or widget
An AI model reads the message and returns urgency (Emergency / Standard / Spam), job type, confidence, and a drafted personalized reply
A switch routes each lead to the correct response branch automatically
AI provider is swappable (Gemini → Claude → GPT) without rebuilding the workflow
Verified
Live test: a "water pouring through my kitchen ceiling" message was classified Emergency, got a personalized dispatch reply, and hit the emergency branch in under 2 seconds.
The edgeTriage over unstructured text — the workflow form-based tools can't build.
Documented portfolio build · ~$0.001 per lead at production scale · full write-up on request
Real roof geometry and material-aware pricing behind a web form — an honest ballpark on the spot.
The workflow itself — n8n workflow canvas
The problem
A homeowner fills out a "get a quote" form, then waits hours or days while someone works up an estimate — and many book whichever contractor answers first. Manual estimating also eats staff time and produces inconsistent numbers.
The system
The homeowner submits dimensions, pitch, material, and number of stories
Input is validated — broken or junk quotes never go out
Calculates the true sloped-roof area, then factors material cost, cutting waste, and multi-story difficulty
Returns a low-to-high range instantly, framed clearly as a starting estimate pending inspection
What changes
An instant answer captures the homeowner while they're still interested
Every quote runs the same logic, tuned to the contractor's real pricing — consistent and defensible
Honest ranges set expectations without over-promising
The speedBallpark price in under a second, any hour of the day.
Documented portfolio build · pricing model configurable per contractor · full write-up on request
Instant quotes and coverage answers in plain English — no agent on the line.
The workflow itself — n8n workflow canvas
The problem
Prospects arrive with two questions — "what will this cost me?" and "how does this work?" — at all hours. A static FAQ can't answer "how much for $500k of term life if I'm 35 and don't smoke?", so the warm lead sits cold overnight waiting for a callback.
The system
Reads each plain-English question and decides at runtime whether it's a pricing or coverage question — agent-style routing, not a hard-coded decision tree
Pricing: extracts age, coverage amount, and risk factors, then computes from controlled pricing logic — the AI rewords the output, it never invents the number
Coverage: pulls the correct explanation from a structured knowledge base
Auto-retries transient API failures; anything out of scope is acknowledged and routed to a human
What changes
A real answer at 11pm on a Sunday, not a form confirmation and a wait
Identical inputs always produce identical, defensible quotes — clearly labelled as estimates pending underwriting
One interface handles both question types, escalating gracefully
The speedConversational answer in under 3 seconds, around the clock.
Documented portfolio build · rate logic configurable per carrier · full write-up on request
A scheduled scraper that builds a qualified, scored prospect list automatically — every week.
The workflow itself — n8n workflow canvas
The problem
Teams that sell to local businesses spend hours hunting prospects — Googling one area at a time, eyeballing which businesses have a fixable weakness worth pitching. The people who should be closing are building lists by hand.
The system
Targets a niche + area from one settings panel — retargetable to any vertical or region in seconds
Infrastructure-level monitoring: know a workflow broke before the client does.
The workflow itself — n8n workflow canvas
The problem
Every client workflow is a silent promise. When one breaks, there's no default mechanism to tell you — the first sign of failure is often the client asking why nothing happened. Manual log-checking doesn't scale past two or three clients.
The system
One handler running at the infrastructure level, wired to every workflow on the instance
Any failure → a structured packet: which workflow failed, the error message, and a direct link to the failed execution
A clear alert email lands within seconds of the failure
Every new client workflow inherits the coverage automatically — there's no per-workflow alerting to forget
What changes
Errors surface immediately, not on the next manual log review
The alert link opens the exact failed run — no searching, no guessing
Clients see problems fixed fast; they never see the monitoring layer
The edgeNo silent failures, ever — alert in seconds, straight to the broken node.
Running live on my production n8n instance · full write-up on request
Build 015 · Zapier · Roofing / Contractors
Speed-to-Lead Capture & Routing
Every inbound lead answered in under a minute — with the owner's attention routed to the jobs that pay.
The workflow itself — Zapier workflow canvas
The problem
When the crew is on a roof, the phone goes unanswered — and the first company to call back wins the job. With leads costing well over $100 each, two missed calls a week is roughly $52,000 a year in lost work.
The system
Every quote request filed in the CRM the moment it arrives — duplicate-proof
Personalized confirmation to the lead within 60 seconds, so they stop dialing competitors
Source tracking: every lead logged to a dashboard showing which marketing channels actually produce jobs
Smart owner alerts sized by job value: replacement/repair → "call within 5 minutes"; routine inspection → "follow up today"
What changes
Under-60-second response, 24/7 — zero leads lost to "I forgot to call back"
High-value work always gets instant attention; routine requests queue without noise
A live view of which lead sources actually convert
Why it mattersTwo missed calls a week ≈ $52,000 a year walking to competitors.
Documented portfolio build · full write-up with workflow screenshots on request
Build 016 · Zapier · Roofing / Contractors
Quote Follow-Up Engine
Three touches over seven days — smart enough to stop itself the moment the quote is signed.
The workflow itself — Zapier workflow canvas
The problem
Most contractors send one quote and wait; quotes go cold in 48–72 hours. The harder problem: automated follow-up must stop the instant a quote is accepted — a "just checking in" email to a client who already signed erodes trust.
The system
Two connected automations sharing a live status flag in a key-value store
Quote marked won → the listener writes "accepted" to the flag
Day 1: immediate follow-up SMS · Day 3: financing-options email · Day 7: a personal-call task for the owner
Before every touch, the engine reads the flag — if the quote was accepted at any point, the sequence stops silently
What changes
Every quote gets exactly 3 strategically spaced touches without the owner remembering anything
No awkward double-contact after signing
The day-7 call list contains only the leads actually worth a personal call
Why it mattersA 3-touch sequence closes 3× more quotes than a single send.
Documented portfolio build · full write-up with workflow screenshots on request
Build 017 · Zapier · Home services
After-Hours Emergency Capture
One intake that knows the difference between 3 AM and 3 PM — and between a flood and a faucet.
The workflow itself — Zapier workflow canvas
The problem
Most emergency HVAC and plumbing calls come in after hours — and they're worth 2–3× a normal job. After hours, the call hits voicemail and the homeowner with a burst pipe dials the next company. An emergency needs to wake someone up; a routine request just needs a friendly callback promise.
The system
Every request logged the moment it arrives — always, regardless of hour
After hours: the on-call team is alerted instantly and the customer gets an immediate acknowledgment
Genuine emergencies (no heat, flooding, gas smell) trigger a direct dispatch to the on-call technician
Business hours: a warm confirmation that the team will call back the same day
What changes
The first-responder advantage, automated — high-value emergency jobs reach a real person in seconds
The most stressed customer (3 AM, water rising) gets reassurance fastest
Runs 24/7 with zero staffing
Why it mattersEmergency jobs run 2–3× normal value — and go to whoever answers.
Documented portfolio build · full write-up with workflow screenshots on request
Build 018 · Zapier · Home services
Post-Job Review Request
Turns every completed job into a Google review ask — inside the window that actually converts.
The workflow itself — Zapier workflow canvas
The problem
72% of customers will leave a review if asked — only 14% ever are. A contractor with 50+ reviews dominates the map pack while competitors with 5 sit invisible. And the window is narrow: response rates drop sharply after 48 hours.
The system
Detects job completion automatically within minutes
Waits 2 hours — customer settled, experience still fresh — then sends a personalized ask referencing their name and the work done
A shorter, softer 24-hour nudge for customers who saw it and forgot
Delivery failure → instant owner alert, so no job slips through silently
What changes
Every completed job asked, within 2 hours, automatically
Zero manual effort from the contractor or office
Reviews compound into local search ranking — the cheapest marketing there is
Why it matters72% will review if asked. Only 14% are. That gap is free ranking.
Documented portfolio build · full write-up with workflow screenshots on request
Build 019 · Zapier · Med spas
No-Show Recovery & Rebooking
Reacts the instant an appointment is missed — and never nags someone who cancelled on purpose.
The workflow itself — Zapier workflow canvas
The problem
Med spas lose 15–30% of appointments to no-shows at $500–$700 each — and most do nothing when it happens. A no-show usually isn't a lost client; they forgot. But you have to tell a genuine no-show apart from a deliberate cancellation, or you end up pestering people who already told you they can't make it.
The system
Reacts the instant the booking system marks an appointment — no polling delay
Smart filtering: ignores completed and confirmed appointments; acts only on no-shows and cancellations
No-show → a warm, no-guilt rebooking offer goes out immediately, while the open slot is fresh
The client's record is tagged, building a no-show history the spa can act on (e.g. deposits for repeat offenders)
Deliberate cancellations are quietly logged — and never sent a rebooking message
What changes
Automatic revenue recovery on appointments that would have just evaporated
Zero risk of nagging a client who already cancelled
Why it mattersEvery rescued no-show is a $500–$700 appointment back on the books.
Documented portfolio build · full write-up with workflow screenshots on request
Build 020 · Zapier · Med spas
Treatment-Clock Rebooking
A per-client timer that fires the rebooking nudge the day results start to fade.
The workflow itself — Zapier workflow canvas
The problem
Aesthetic treatments run on a predictable clock — Botox ~3 months, filler ~9. The ideal rebooking day is different for every client, and no front desk can track hundreds of individual clocks by hand. Industry-wide, more than half of rebooking windows slip by.
The system
Each client's rebooking date is calculated automatically from their last visit plus their treatment interval
Every morning, the engine scans the roster for anyone due that day
Each due client gets a personalized message — their name, their treatment, their timing — from one smart template
Days with no one due send nothing: no spam, no false nudges
What changes
Every client contacted on exactly the right day — the single biggest lever on repeat revenue in aesthetics
Zero front-desk effort; 50 clients or 500, same single workflow
Adding a new treatment type never means rebuilding the automation
Why it mattersOver half of rebooking windows are normally missed — this catches every one.
Documented portfolio build · full write-up with workflow screenshots on request
Build 021 · Zapier · Insurance agencies
Quote → CRM + Instant Follow-Up
Files the lead, opens the deal, and sends the right email — before the agent even sees the alert.
The workflow itself — Zapier workflow canvas
The problem
Agents can't watch the quote form around the clock — and 40% of requests go cold within 15 minutes if no one responds. Meanwhile there's no CRM record, no pipeline deal, no follow-up scheduled. Every slipped quote is a policy that never got written.
The system
Instant CRM entry the moment they submit — duplicate-proof on repeat submissions
A deal opens automatically in the pipeline, tied to the contact, with estimated coverage pre-filled
Every submission appended to a tracking dashboard: source, policy type, coverage amount
Agent alerted within seconds — who submitted, what they need, and a 15-minute follow-up prompt
Confirmation email routed by policy type: personal-lines copy for residential, business copy for commercial
What changes
Every quote request answered in under 60 seconds, 24/7
Every lead in the CRM with a live deal from the moment they submit
Why it matters40% of quote requests die in 15 minutes of silence — this answers in 60 seconds.
Documented portfolio build · full write-up with workflow screenshots on request
Build 022 · Zapier · Insurance agencies
Policy Renewal Sequence
A disciplined 60 / 30 / 7-day cadence across the whole book — escalating exactly when it should.
The workflow itself — Zapier workflow canvas
The problem
Renewals are the most predictable revenue an agency has — every policy has a known date — yet they're chased manually, so windows slip. Across a book, that's 10–15% annual attrition: clients who drifted away because no one reminded them at the right moment.
The system
A daily engine scans the entire policy book; each policy's reminder dates are computed from its renewal date
60 days out → friendly heads-up email
30 days out → email + SMS reminder
7 days out → email + SMS + a call task for the agent to phone the client personally
Quiet on days nothing is due; escalating exactly when each policy needs it
What changes
No missed renewal windows — the whole book worked on the right days, automatically
The agent's personal time goes only to the final, at-risk tier
50 policies or 5,000 — same engine, every morning
Why it mattersAttacks the 10–15% of the book that quietly lapses every year.
Documented portfolio build · full write-up with workflow screenshots on request
