Build the hail campaign in February and leave it paused
The worst possible time to write a storm campaign is the week you need it running.
After a significant hail event, the auction crowds up fast. Writing ad copy, building a landing page, setting negatives and configuring call tracking while the office phone is ringing produces bad versions of all four.
Do it in the quiet months instead. A separate campaign, its own ad groups, its own landing page, a daily budget you have already agreed on, and an exclusion list you have already argued about. Then pause it.
Write the trigger rule down as well. Who turns it on, on what evidence, at what budget, and what has to be true about crew capacity before it goes live. A campaign that runs while you are three weeks behind on installs buys somebody else's roof.
Keep the same discipline on the way out. Storm campaigns should have an agreed condition for pausing, not a vague sense that things have calmed down.
The addresses your crews will not reach in January belong in the exclusions
Geography for a roofing company is drive time and access, not a radius drawn around the yard.
A circle centered on your shop will happily spend money in places a crew cannot service at a sensible cost. Douglas County down toward Castle Rock is a different day than a job in Adams County, and both are different from foothills access in winter.
Split campaigns by the zones your dispatcher already thinks in, and give each one its own budget. If the numbers say one zone costs more to win work in, that is useful information rather than a problem to average away.
Set location targeting to people in your area rather than people interested in it. Hail draws attention from out of state, and some of that attention is not a homeowner.
Then write the exclusions honestly. Streets you will not climb in February, mountain addresses you would decline anyway, and any zone where you already have more work than crews.
Impact resistant shingle searches cost more and close better than roof repair
Intent varies enormously by phrase, and blending it all into one cost per lead hides the only thing worth knowing.
Repair terms are cheap and busy. Some of them are real work, many are small jobs, and a share are people who want a phone opinion. That traffic can be worth buying, but not from the same budget as replacement.
Material and upgrade searches behave differently. Somebody comparing impact rated laminate against a metal roof, or searching a manufacturer line by name, is generally further along and closer to a real decision.
Give each job type its own campaign, its own landing page and its own target. Repair, full replacement, material upgrades, and commercial or low slope work if you sell it.
Brand terms deserve their own line as well, both yours and the competitors homeowners are comparing you against after a canvasser leaves a card.
Solar shoppers click roofing ads here, and they never book a tear off
In a Denver roofing account, the exclusion list does more work than the bid adjustments.
Solar is everywhere in this market and the search language overlaps constantly with roofing. Decide first whether you sell panel removal and reinstallation for a reroof. If you do, build a small campaign for it. If you do not, exclude the whole family and stop paying for it.
Other steady drains: hiring and apprenticeship searches, DIY tutorials, supply and material pricing lookups, gutter cleaning, and anything aimed at training or certification.
Read search terms weekly in normal months and daily during storm season. New waste appears fastest exactly when spend is highest.
Keep adjuster and claims language out of your keywords and your ad copy. Explain roofs, not policies, and let homeowners take insurance questions to their own insurer.
The week after the storm is a better buy than the afternoon of it
The storm cycle has phases, and the most expensive hour is rarely the most productive one.
The afternoon a cell passes, the street fills with canvassers and the searching is chaotic. Tarps, damage photos, curiosity, and a lot of clicks that were never going to become an inspection.
A few days later the picture changes. Homeowners have talked to their insurer, walked the yard, collected a card or two, and started comparing companies. Ads written for that moment do better work than ads written for panic.
Pace accordingly. Daily caps that reflect crew capacity, a rule for shifting budget between zones, and a plan for what happens when one week threatens to eat a quarter of the annual spend.
Between storms, keep a smaller always on campaign alive for replacement and repair. Roofs wear out along the Front Range whether or not it hails, and those months are usually cheaper to buy.
Blend the storm months with the quiet months and the reporting stops meaning anything
A single averaged cost per lead across a hail year describes nothing that happened.
Report storm driven work and planned replacement separately. They have different close rates, different job sizes and different acquisition costs, and averaging them together tells you nothing you can act on.
Push signed jobs back into the ad platform where your systems allow it. Optimizing toward form fills teaches the algorithm to find people who fill in forms, which is not the same population as people who sign a contract.
Tag every inquiry with its county and its zone. Denver is a consolidated city and county while its neighbors run their own show, and cost to win work genuinely differs across a modest drive.
Record calls where permitted, and listen to a handful every week. Most paid search problems in this trade are intake problems wearing a media costume.
Questions we actually get
- Should we run paid search at all if we already get storm work by knocking doors?
- Paid search reaches the homeowner who took a canvasser's card inside and started comparing. It also covers the months when nobody is knocking. Whether it earns its place depends on your capacity and your close rate on inbound work, which is worth measuring before committing a budget rather than after.
- How much budget does a Denver roofing account need?
- There is no honest single answer, and any number quoted without seeing your service area, job mix and crew capacity is guesswork. The more useful framing is how many additional inspections a week your estimators can absorb, then working backward from what those cost in your zones.
- Can we advertise on hail damage terms without stepping into insurance territory?
- You can advertise roof inspection and replacement services. Avoid copy that discusses deductibles, promises a claim outcome, or offers to handle a claim. Point people to their own insurer, and in most cases to their own counsel if a dispute arises. It is also worth confirming your language with your own advisors, since the rules around this are taken seriously.
- Do Local Services Ads change the picture here?
- They sit in a different part of the results page and are worth evaluating separately from search campaigns, including the verification requirements and how disputes over charged leads are handled. Treat them as an additional channel with their own economics rather than a replacement for a structured search account.
- What number should we ask an agency to report on?
- Cost per signed roof, split by zone and by job type, with storm driven work reported separately from planned replacement. Leads and clicks are useful for diagnosis, but they are not results. If the reporting cannot connect a signed contract back to a campaign, the account is being managed blind.
