A trade that does not spend evenly
Most advertising thinking assumes something close to steady demand: set a monthly budget, run it evenly, and let the campaign tick along. Roofing does not behave that way, and pretending it does is the most common way a roofing campaign wastes money. Demand for roofing work is weather-driven and spiky. Long stretches go by with ordinary, plannable enquiry volume, and then a run of bad weather compresses what would have been weeks of enquiries into a matter of days.
A campaign spending the same amount every week is wrong for that shape of demand in both directions. Through the quiet stretch, you are bidding for a smaller pool of people who are mostly just starting to think about a re-roof, at whatever the going rate is that week. Then the week arrives when search intent genuinely explodes — everyone whose roof has just failed is typing the same handful of searches at once — and a flat weekly budget has no extra room to spend into it. You end up underspending in the exact week the enquiries were worth the most, and overspending through the stretch either side where they were worth the least. That is not a rounding error. It is the entire logic of the campaign working against the shape of the trade it is meant to be serving.
Two businesses, two kinds of ad enquiry
Roofing splits cleanly into two kinds of work, and a campaign that treats them as one audience will get the tone, the channel and the follow-up wrong for at least one of them.
The urgent side is leak and storm damage. Someone has water coming through a ceiling, or has just watched a section of roofing peel back in the wind, and they are not comparing options in any considered sense. They want to know who can come and how soon. This customer is not weighing up quotes over a week — they are ringing down a list, and the first roofer who gives them a real answer tends to get the inspection booked. Search fits this half of the business closely, because the customer already knows they have a problem and is actively looking for someone to fix it. You are not creating the need; you are being found at the moment it exists.
The planned side is replacement or re-roofing — a job someone has been meaning to get quoted, is budgeting for over a season, and is often genuinely comparing more than one roofer before committing. This customer has not necessarily gone looking yet. A homeowner who knows their roof is getting old but has not actively searched for a roofer is exactly the kind of interest Meta is built to interrupt: a well-shot before-and-after in a feed can start a decision process that a search campaign would never have reached, because the customer was not typing anything into Google yet. This side of the business is also, more often than the urgent side, insurance-adjacent — a claim is sometimes part of how the job gets funded, which is worth knowing going in only because it means the customer may be dealing with an insurer’s own process alongside yours, not because it changes anything about how you should sell the job.
Treating these as one audience, with one message and one channel, means doing badly by whichever half gets the leftover attention. A search ad written for someone with an active leak reads as pushy and irrelevant to someone quietly researching a re-roof for next year. A relaxed, visual Meta campaign built for the planned side does nothing for someone standing under a bucket right now. They need to be told apart from the moment the budget gets set, not sorted out once the enquiries start arriving.
The cruel part
Here is the timing problem that makes roofing the hardest trade on this site to advertise sensibly, and it is worth sitting with properly because it is not obvious until you have lived through it.
The week after bad weather is exactly when a roofing campaign performs best. Search intent goes up, because more people than usual have a genuine, urgent reason to look for a roofer. It is also the week competitors bid hardest for the same searches, because everyone running a campaign can see the same spike you can. And it is also, unavoidably, the week you are flat out on emergency callouts — tarping, temporary patches, inspections on properties already on your books — with the least spare capacity all year to pick up a phone.
That is the cruel part. The moment your ads work best is the moment you are least equipped to convert what they bring in. A search campaign that has been switched on and budgeted for ahead of a spike is putting your business in front of exactly the people who need you most, at exactly the moment you have the least room to answer them. Every enquiry that arrives in that window and goes unanswered is not a neutral miss. It is money spent twice over — once on the click that generated the enquiry, and again on whatever it costs to lose a customer who was, at that moment, ready to book.
The general case for reply speed covers what counts as fast enough on an ordinary day. During a genuine weather spike, that guidance is a floor rather than a target — the gap between a fast reply and a slow one during a storm week is wider than almost any other week of the year, because the customer’s patience has already run out before they picked up the phone. What actually happens to roofing enquiries in the week after bad weather goes into that mechanism in full; the point worth carrying into a campaign specifically is that spend and answering capacity need to move together, not spend running high while answering capacity is at its lowest point of the year.
What that means for how a campaign gets built
None of this means avoiding the urgent side of the trade, or refusing to advertise into weather spikes because they are hard to staff. It means building the campaign with the shape of the demand in mind rather than against it. Search budget for the urgent side should have room to flex upward when conditions suggest demand is about to spike, rather than sitting fixed at a figure that made sense on an ordinary week. Meta budget for the planned side can run on a steadier rhythm, because a re-roof decision does not move with the weather the way a leak call does — it is closer to a normal considered purchase, and remarketing to someone who has already shown interest fits that longer decision naturally.
The other half of getting this right has nothing to do with the ad platform at all. A campaign that spends more into a spike is only a good idea if there is somewhere reliable for the enquiries it generates to go — some way for a call to get a fast reply even when every hand on the team is already up a ladder. Without that, spending harder into the exact week you are least available just multiplies the number of paid enquiries going unanswered.
Where the gate matters most
This is why we do not run campaigns for a roofing business until the front desk is in place. Of all the trades on this site, roofing has the sharpest mismatch between when demand is highest and when capacity to answer it is lowest, which means it also has the most to lose from buying enquiries into a system that cannot catch them. A campaign switched on before enquiry handling is sorted does not create more roofing business. It creates more evidence of the same leak, paid for at the worst possible moment — the week a customer with a genuine leak rang, nobody picked up in time, and they hired whoever answered next.
The roofers page sets out how the front desk itself works — capturing enquiries, replying fast, and following up on the inspections and quotes that come out of both the urgent and the planned side of the business. Once that is in place, paid campaigns stop being a bet against your own capacity and start doing what they are meant to: putting your business in front of the right person, on the right channel, at the moment they need a roofer — with somewhere for that enquiry to land that will not let it go quiet.