An Arizona service business running a flat monthly ad budget is making a specific, expensive mistake twelve times a year. Half the year it underspends into demand it could have captured. The other half it overspends into demand that no longer exists.
Here is how to run a Valley account properly.
The budget should not be flat
Start here because it is the largest single error.
Arizona demand for cooling, pools, and outdoor services is not seasonal in the gentle sense. It steps. A March budget and a July budget should not be the same number, and in most accounts they are because the budget was set once and never revisited.
The correct shape for most Valley service businesses:
January through March: lower spend, weighted toward planning-intent and maintenance-plan keywords rather than emergency terms. This is where you fill the slow months and build the customer base that reduces summer chaos.
April: ramp. Pre-season maintenance demand is real and it is far cheaper to capture than July emergency clicks.
May through September: peak spend. This is where the year is made. Underspending here to protect an annual budget is the most expensive kind of discipline.
October and November: shift toward seasonal-resident and reopening intent, which behaves differently and often converts at better margins.
December: lowest.
If your account has run the same daily budget since last year, you have almost certainly left summer revenue uncaptured and paid for January clicks you did not need.
Dayparting: your peak is after you close
Arizona's daily demand curve for emergency service peaks between roughly 4pm and 8pm. Most office schedules end at 5.
Two implications for the ad account.
Do not cut ads at 5pm. A lot of Valley accounts daypart to office hours to avoid paying for clicks nobody can answer. That is solving the wrong problem — it removes you from the highest-intent window of the day. Fix the answering, then bid into the evening.
Bid up the evening deliberately. If your data shows evening clicks convert better, and in Arizona they usually do, that should be reflected in bid adjustments rather than left to chance.
The prerequisite is that someone or something actually answers after 5. If nothing does, evening ad spend is genuinely wasted, but the fix is coverage, not retreat.
Build a monsoon preset before monsoon season
Monsoon storms produce near-vertical demand curves with no warning. A microburst at 8:30pm generates its search volume at 8:45pm.
By the time a human notices, raises budgets, and adjusts targeting, the window has largely closed. Automated bidding helps but learns too slowly for a two-day event.
What to build in advance: a saved configuration with elevated budget caps, tightened geographic targeting to the storm-affected corridor, and ad copy written for storm damage rather than routine service. Activate it the evening a storm hits and revert it a few days later.
This is a fifteen-minute setup task in the spring that captures the most valuable clicks of the monsoon season.
Feed the algorithm real outcomes
Automated bidding generally beats manual management for Valley service businesses, but only if the conversion data means something.
Most accounts count form fills and calls over thirty seconds. That set includes wrong-service-area inquiries, price shoppers, existing customers calling about scheduling, and vendors. The algorithm optimizes toward more of the whole mixture.
Send back booked jobs with revenue values from your CRM instead. A system optimizing toward a full system replacement behaves very differently from one optimizing toward an undifferentiated lead count, and in a market where your ticket range spans from a $99 service call to a $14,000 replacement, that difference is enormous.
Geography is not a radius
A Phoenix-based contractor targeting a thirty-mile radius is bidding on Buckeye, Queen Creek, Anthem, and Apache Junction, wildly different drive times and therefore wildly different job economics at the same ticket.
Target named locations rather than radius. Use presence-only targeting so you are not paying for people merely showing interest in the area. Apply bid adjustments that reflect real transit economics: a job forty-five minutes out is worth less to you than one fifteen minutes out even at identical revenue.
Copy that is not interchangeable
Every Valley HVAC ad says fast, reliable, licensed, insured, free estimates. None of it differentiates.
What works is specific and verifiable. A real response window. A named service area. A concrete guarantee. Same-day availability stated honestly, including when you do not have it, a caller told "next available is tomorrow 8am" converts far better than one promised same-day and then rescheduled.
During peak season, honesty about lead time in the ad copy actually improves account performance, because it pre-qualifies out the people who will not wait and stops you paying for clicks that were never going to close.
Performance Max with guardrails
Performance Max will spend a local service budget aggressively across placements that generate volume and no jobs.
If you run it: brand exclusions, account-level negative keyword lists, asset groups split by service line rather than lumped, and a monthly review of the placement report. Unconstrained, a meaningful share of Valley Performance Max spend routinely lands somewhere useless.
What to measure
Cost per lead is the metric that makes bad accounts look good. Look at cost per booked job and return on ad spend using real revenue.
Then look at both by month. An annual average in Arizona is close to meaningless, it blends a July where you could not answer the phone fast enough with a January where clicks were cheap and intent was low. The story is entirely in the monthly detail.
The checklist
- Budget varies by month, deliberately, with a documented seasonal plan
- Evening hours are covered by ads and by someone who answers
- Monsoon preset built and saved before June
- Conversions defined as booked jobs with revenue values fed from the CRM
- Named-location targeting, presence-only, with drive-time bid adjustments
- Landing pages match the ad, emergency ads to emergency pages
- Placement and search-term reports reviewed monthly
Most Valley accounts are missing at least four of these. Each one is an afternoon of work and each one compounds through the season that makes your year.