Roof repair leads Sydney: 35 enquiries at $116.85

A plumbing account is not automatically a roofing account

Aussie Plumbing had a working Sydney search account before any of this started. Plumbing, drains, the emergency work that pays for a trade in this city — all of it already running, already producing enquiries at a price the business had decided it could live with. Roof repairs looked, from the outside, like the easy next step. Same brand, same vans, same suburbs, same phone.

It is not the easy next step, and the reason is worth stating plainly. A roof leak and a burst pipe feel like the same emergency to the person searching, but they are bought differently. The roofing auction has specialist restoration firms in it who do nothing else. The searches carry a lot of noise a plumbing account has never had to deal with — roof racks, car roofs, sunroofs, people looking for a job on a roofing crew, people pricing tiles at a hardware store. And the intent splits in a way plumbing rarely does: half the market wants somebody there today because water is coming through a ceiling, the other half is pricing a restoration they might get around to next autumn.

None of that is visible in the plumbing account's history, so none of it could be borrowed. The roofing side had to learn it the way every new campaign does — by spending money and being read closely by somebody while it did.

What the first five days on air actually told us

The campaign went live and ran for five days. It spent $639.50, drew 35 clicks and produced 4 enquiries, which works out at $159.88 each.

We want to be careful about that figure, because it is the kind of number that gets quoted as a starting point and then used as the base of a percentage. Four enquiries is not a baseline. It is a ramp. Off four, a single unusually cheap week or one expensive one moves the average by thirty or forty dollars, and anybody who tells you otherwise is selling the drop rather than reporting it.

What those five days were genuinely good for was the search terms. Somebody sat down and read what people had actually typed to reach the ads, which at 35 clicks takes about twenty minutes and is the single most useful twenty minutes in a new campaign's life. The reading is what produced the decision, and the decision was not "adjust the bids".

The campaign was not underperforming. It was aimed at the wrong words.

Five days off air

We took it down. For five days the roofing campaign served nothing at all, which is an uncomfortable thing to do to a client who has just started paying for it, and the right call anyway.

What went on in that gap was ordinary craft work, done by people at desks. The ad groups were pulled apart and rebuilt around what the search terms showed people were actually looking for — repairs, leaks, restoration, repointing, broken tiles, metal and Colorbond — rather than around the trade's own word for it. Somebody wrote fresh ads for each of those, because an ad written about roof plumbing does not speak to a person who typed "cracked roof tiles". And a long list of exclusions was written by hand, one at a time, to keep the account away from roof racks, sunroofs, hardware pricing and job seekers.

That last part is unglamorous and it is most of the job. There is no clever way to do it. Somebody reads the terms, decides which ones are a person with a wet ceiling and which ones are a person with a Corolla, and writes the second group out of the account.

The fourteen days after

• 31 enquiries in 14 days — from 287 clicks and 5,233 impressions
• $111.30 per enquiry, against $159.88 across the five days before the rebuild
• 10.80% of clicks became an enquiry, against 11.43% in the opening stretch
• $3,450.35 spent in those 14 days, against $639.50 in the five before
• Roughly 2.2 enquiries a day, against 0.8 a day before

The number worth holding onto there is not the cost per enquiry. It is the conversion rate, which barely moved — 11.43% before, 10.80% after. That is the figure that tells you what actually happened.

If the rebuild had simply found a cheaper, softer corner of the market, the conversion rate would have fallen away as volume climbed. It did not. The account went from 0.8 enquiries a day to 2.2, spent nearly twice as much per day to do it, and kept converting at roughly the same rate — which says it found more of the same kind of person rather than a lot more of a worse one.

Across the whole 25 days

Taken end to end rather than split in two, the campaign spent $4,089.85, served 5,595 impressions, drew 322 clicks at a 5.76% click-through rate, and recorded 35 enquiries at a blended $116.85 each. 10.87% of clicks became an enquiry.

Twenty-eight of the 35 came in as phone calls and 7 as submitted forms, which is the split you would expect and worth planning around. Two enquiries a day arriving as calls is a rostering question before it is a marketing one — a roof leak customer who reaches a voicemail rings the next roofer on the list, and no amount of advertising fixes that.

The device split is less predictable than the plumbing side of the same account. Mobile produced 21 of the 35 enquiries off 222 clicks, but desktop produced 13 off only 89 — a materially better rate per click. That is a small sample and we are not going to build a theory on it, but it is consistent with the half of this market that is pricing a restoration rather than reacting to an emergency, and that half sits at a desk.

What we are not claiming

Three things, stated before anybody else has to ask.

This is 25 days. It is not a trend, it is not a season, and roofing is a trade where the weather decides a good deal of the demand. A wet fortnight in Sydney will flatter any roofing account and a dry one will punish it, and 25 days is nowhere near enough to see through that.

The $159.88 opening figure rests on 4 enquiries. We have quoted it because it is what the account did, and we have put the drop to $111.30 next to it because that is also what the account did. We are not going to render it as a headline percentage, because a percentage calculated off four events reads as far more solid than it is.

And the roofing campaign is not the account. It sits alongside the plumbing campaigns in the same Aussie Plumbing account, which has its own longer history and its own figures. This page is about the roofing side only, over one measured window, reported without dates on purpose.

Adding a second trade to a working account?

The mistake we see most often is treating it as an extension rather than as a new campaign that happens to share a logo. The existing account's history does not transfer, the search noise is different, and the first fortnight will look worse than the business expects because it is genuinely starting from nothing. What it needs is not a bigger budget. It needs somebody to sit down and read what people typed, then act on it quickly enough that the learning does not cost a whole quarter. If you are putting a second trade in front of a Sydney audience, talk to us before you switch it on.