Two shops, one state, four thousand feet apart. The ad account thought they had the same weather

Who this is

1st Choice sells plumbing, air conditioning and insulation out of two shops: Phoenix and Prescott. About a hundred miles apart on the map, and about four thousand feet apart vertically.

That second number is the one that matters. Phoenix sits around eleven hundred feet and does not freeze. Prescott sits around fifty-four hundred, gets snow most winters, and has a heating season that Phoenix simply does not have.

One account, two calendars

The advertising did not know about the elevation. One set of campaigns covered both shops, which means one budget, one seasonal shape, and one idea of what people are searching for in January.

In Phoenix in January, people are searching for very little. So the account throttled down for winter — correctly, for Phoenix. Prescott's heating season was throttled down with it, in the months it needed the money most.

The reverse happened in July. Phoenix at capacity pulled the budget, and Prescott — where July is a perfectly normal working month — got whatever was left.

The band is Prescott's strongest stretch of the year. It is also the stretch in which the shared budget was at its lowest, every year, on purpose.

The money that went nowhere

Before any of the seasonal work, there was a simpler problem. A hundred miles of mostly empty Arizona separates the two shops, and the combined targeting covered a good deal of it.

Black Canyon City, Cordes Lakes, the stretch of the I-17 in between. Real places with real households, and nearly an hour from either shop. The account had been buying them for years because they fall inside a radius drawn around two points at once.

What we did not expect

We assumed the Prescott side would need a smaller version of the Phoenix campaigns. It needed a different one.

Prescott is a much older market in the demographic sense, with a housing stock full of systems that predate anything being sold in a Phoenix suburb. The searches are not the same searches, the objections are not the same objections, and the Phoenix landing page — written for a metro of newer builds — was answering questions the Prescott caller had not asked.

That page had been live for two years. It ranked. It just was not talking to anybody in Yavapai County.

What came back

Prescott stopped vanishing every November. The cost per booked visit came down across both locations, and most of that is the money no longer going to the hundred miles in between.

The insulation side is where this gets genuinely interesting, and it is also where we are least confident. Insulation sells for different reasons at two different elevations — heat exclusion in Phoenix, heat retention in Prescott — and those are two different arguments that had been sharing one advertisement.

The overlap has to be decided by a person

Two shops always produce a middle, and the middle cannot be left to the platform. If both locations are eligible for the same household, the same company enters the same auction twice and pays the difference for the privilege.

We assigned every contested area to one shop on drive time, wrote it down, and stopped revisiting it. It is not a sophisticated decision. It is just one that has to be made explicitly, because the default is to make it badly.

What this does not prove

• The company does not publish a licence number in its main footer, so unlike the other cases here we cannot point at a registry entry as the cheapest proof this is a real contractor. We have asked.
• Prescott is the smaller of the two markets by some distance, so its percentages move on a small base.
• A mild Prescott winter and a hard one are very different years, and the two winters in this comparison were not equivalent. Some of the improvement is weather.
• Cutting the corridor between the shops improved the cost per booked visit and also removed customers. We have not shown that the corridor is unprofitable at every budget, only that it was the worst use of this one.
• The insulation split by elevation is the newest part of this work and has the least evidence behind it. We believe it and we cannot yet prove it.

What to take from this

1. Check the elevation before you copy the campaignTwo locations in one state can be on completely different heating calendars, and a budget shaped for one of them starves the other in its best months.
2. Give each location its own seasonal curveA shared budget throttles by the dominant market's calendar, so the smaller market is cut back in exactly the months it should be scaling.
3. Do not let a radius bridge the gap between two shopsThe empty ground between locations falls inside both circles and gets bought twice, at a drive time neither shop can service profitably.
4. Write the second location its own landing pageAn older market with older housing stock asks different questions, and a page that ranks is not the same thing as a page that answers them.

Questions people ask about this one

Can two locations share one advertising account?

They can share an account. They should not share a budget, a service area or a seasonal schedule. The account is an administrative container; the budget is a decision about which market gets to spend in which month. Merging those two things is what causes the smaller location to be quietly defunded during its own peak season.

How do you decide who owns the area between two shops?

On drive time from each address, not on distance or on postcode. Then you write it down and stop changing it. The overlap is not a difficult judgement, but it has to be an explicit one, because the default behaviour is for both locations to bid on the same household and for you to pay the difference.

Is a hundred miles too far apart for one company?

Not at all — it is two businesses that happen to share an owner, a brand and a back office. The mistake is treating them as one business with a wide radius. Nearly an hour of empty road separates the two service areas here, and the households living in it are not profitably reachable from either shop.

Does insulation really sell differently at different elevations?

We believe so, and this is the part of the case with the least evidence behind it. In Phoenix the argument is keeping heat out of a house in July. In Prescott it is keeping heat in during a winter with snow on the ground. Same product, opposite pitch, and the single advertisement they had been sharing could only make one of those arguments well.