A roofing company turned storm-season form fills into scheduled estimates automatically

Representative examplesEvery case study in this library is an illustrative composite of the kind of engagement we deliver — written to show our method and standards, not to name clients.

Where they started

Weather sets the calendar for this regional roofing contractor: hail or wind converts a slow month into a flood of estimate requests within days, often from homeowners who have never hired a roofer and are calling three companies at once. Crews are scheduled around conditions across a three-county area, and a two-office staff answers everything. Estimates are free, but drive time is real — the estimators' calendars are the company's true capacity. Requests used to arrive by phone, form, and referral, and were worked in whatever order someone got to them.

What it was costing

Storm season produced a surge of estimate requests that the office answered in whatever order voicemails arrived. Homeowners waited days, signed with whoever called first, and the company's own close rate fell during its busiest weeks.

What they could see

• During storm weeks, form submissions waited behind voicemails until somebody had a hand free.
• Homeowners described calling three roofers and hearing back from this company last, or not at all.
• The estimators' days were assembled by the office around whatever sounded most urgent that morning.
• Insurance-claim callers heard a price-first script and hung up confused about the deductible process.
• By late season the office had stopped logging requests they knew they could not call same-day.

The constraints we worked inside

• Storm damage makes the buyer anxious — response speed and clarity outrank polish.
• Estimators' calendars filled in real time; the funnel had to book into actual availability.
• Insurance-claim customers needed different framing than cash customers — one page for both failed both.

What had been tried before

The company hired a seasonal office temp to help answer the storm-month surge.

A second person answering in the same order made the queue faster, not smarter; the insurance and cash paths still tangled, and retraining a temp every year cost more than it returned.

An answering service was trialed for overflow calls during peak weeks.

The service could take a message, not an estimate request — homeowners had to repeat themselves to the office later, and the delay stayed exactly where it had always been.

A scheduling whiteboard with three colors was introduced for the estimators' week.

It worked for one estimator and not the other two, and storm weeks erased it entirely — the board was the first casualty of every surge, and everyone knew it.

What we proposed

Split the funnel by customer situation, because an insurance-claim conversation and a cash-quote conversation do not share a first sentence: separate pages and forms, each answering its own questions — deductible process for claim customers, price and scope for cash. Estimate slots connect to the estimators' real calendars with travel-zone buffers, so bookings are keepable rather than theoretical. Instant confirmation, a preparation checklist, and day-before reminders run as workflows in plain language, and every request lands as a stage-tracked card the office can actually see.

Just as important is what we ruled out, and why:

• A 24/7 call center taking estimate bookings — Storm-damaged homeowners ask specific questions about damage types; a scripted service would book wrong-priority visits the estimators then have to untangle in the field.
• One combined page with a claim-type dropdown — The two situations differ at the first sentence, and a dropdown kept every homeowner on the same page and the same first impression — which is precisely what had failed before.
• Buying surge capacity from third-party lead services — They sell the same storm leads to every roofer in the county, which is exactly the race the company was losing; the fix had to make owned requests faster, not buy more of the shared pile.

How the work ran

Delivered by the growth pod — strategist + automation specialist over 4 weeks, with working increments reviewed with the client every week.

The stack, and the reasoning

GoHighLevel funnels

Storm weeks flood both paths within days, so the claim and cash pages double as surge triage — the office sees which situation each caller is in before dialing back, and the builder shipped both without a web project.

Calendars & booking

Slots are honest about drive time across the three-county service area: the calendar offers only windows an estimator can actually reach from the day's routes, because a missed storm-week estimate is a day the crew never gets back.

Workflows

Confirmation, preparation checklist, and day-before reminders run themselves in plain language for anxious homeowners, so the office stops being the pacing item at nine at night after a storm.

Opportunities & Pipelines

Every request becomes an estimate card carrying the homeowner's own photos of the damage, the storm date, and its stage, so the estimator sees the roof before the driveway and the office sees the surge on one board.

What went wrong

How we worked together

Cadence

A daily 10-minute stand-up during the pilot weeks, then twice weekly; the pipeline board stayed on the office wall screen all day where everyone could see it.

Client side

The office manager owned stage rules and copy tone; the lead estimator owned the calendar buffers; the owner appeared when money or messaging needed a call.

Decisions

The morning stand-up made operational calls on the spot; anything structural waited for the Friday review with the owner, who decided by walking the board.

They provided

Estimator schedules and service zones, examples of real claim and cash conversations, the phone scripts they actually use, and the office's full attention during pilot weeks.

What changed

The headline: median form-to-scheduled-estimate time during the storm season — 2.4 days → 4 hrs, read from Pipeline timestamps. A second check: estimates held (booked and completed) versus prior season at +26%.

Storm season stopped being a triage exercise. Homeowners get a confirmation in minutes that names their situation — claim or cash — and tells them what happens next; the office reports that calls start calmer because the first text did the reassuring. The estimators run routes instead of scavenged schedules, and they hold more of the estimates they drive to. There is no unreachable pile anymore, because every request is a card on the board with a next step, even at the height of a hail week.

The result was read from Pipeline timestamps against the pre-engagement baseline over the stated window, with a guardrail check on estimates held (booked and completed) versus prior season. Where platform-reported numbers and business outcomes differ, this record says which layer it is quoting.

What they own now

• Both funnel paths — claim and cash — with forms wired to the estimators' calendars.
• The travel-zone buffer settings and the rule for updating them each season.
• Workflow templates for confirmation, preparation checklist, and reminders.
• The pipeline stage guide the office trained itself on.
• A storm-season readiness checklist run before each year's first forecast.