Storefront glazing on the Strip is not your job, and the auction cannot tell
The exclusion list is the highest return work in a valley account, and most of it is specific to this city.
Broad window and glass terms in this market pull in a remarkable amount of work you do not do. Commercial storefront and curtain wall queries from the Strip corridor and the resort properties. Auto glass and vehicle window tinting, which in a place this sunny is a large industry of its own. Sun screens, solar film, screen repair, single pane replacement, shower doors, mirrors.
Each of those clicks costs real money and none of them books a measure. Left unmanaged, they can absorb a meaningful share of a small account before anyone opens the search terms report.
Build the negative list before launch, not after the first bad month. Then read the actual search terms weekly for the first stretch and monthly after that. New junk appears constantly as the auction broadens matching on its own.
Also exclude the research intent you cannot serve. Do it yourself sizing, replacement glass only, warranty lookups for a builder's original units, and new construction bid requests unless builder work is a line of business you actually want.
The patio slider deserves its own budget, not a line in the whole-home campaign
Valley housing stock makes the sliding door a distinct product with a distinct buyer, and blending it into whole-home replacement distorts everything downstream.
A homeowner replacing one failed slider to the back patio is a different sale from a homeowner replacing every opening in a two story house. Different price, different urgency, different objection, different close rate.
Blend them and the averages become meaningless. Cost per lead looks fine because the slider queries are cheaper. Cost per booked job looks terrible because the whole-home budget was starved by them.
Run them as separate campaigns with separate budgets and separate landing pages. Entry doors are usually a third. If you serve rental portfolios or property managers across Spring Valley and North Las Vegas, that is a fourth conversation with its own message.
Separation also gives you a lever. When install capacity tightens, you can throttle the small ticket product without touching the campaign that fills the calendar for the next month.
One bid for Spring Valley and Boulder City pays for the drive twice
Geography in this valley is a cost input, not a targeting checkbox.
A radius drawn around a showroom is the laziest setting in the account. It buys clicks from wherever the circle lands, at the same price, regardless of what the trip costs the crew.
Split geography into groups that reflect real operating cost. The core corridor your trucks already run. The outer master plans in Summerlin, Anthem and Centennial Hills. Henderson and North Las Vegas, which are their own cities with their own permit processes. Boulder City if you go, which should carry a lower bid or a minimum job size.
Use location targeting on presence rather than interest, or you will pay for people in another state searching about a vacation property. Then adjust bids by observed performance per group instead of by instinct.
One more thing worth pricing honestly: a measure that requires crossing the valley in the afternoon on the 215 is not the same asset as one ten minutes from the yard, even at the same ticket.
Turning ads off at six is a mistake in a town that never fully closes
A round the clock hospitality economy means a large share of your market is awake and reachable when a standard account has gone dark.
Dayparting rules copied from another market assume everyone works nine to five. Here, a significant number of households run on casino, resort, airport and hospital schedules. Their errand hours are somebody else's night.
Look at when your conversions actually happen before you restrict anything. In our experience the pattern in this valley rarely matches the template, and the cheapest incremental demand often sits in hours competitors have switched off.
Then make sure the offer matches the hour. If nobody answers the phone at two in the morning, do not send that click to a page whose only action is a call button. Send it to a form or a self scheduling page that captures the job while the intent is live.
After hours availability is worth advertising here in a way it is not in most cities, but only claim it if the phone is genuinely covered. An unanswered promise is worse than no promise.
Spend should follow the install calendar, because nobody wants a wall opened in July
Seasonality in the Mojave runs on heat, and the smart pacing curve is not the one a national playbook hands you.
Demand and delivery are not in phase. Interest tends to build as the heat becomes intolerable, while the comfortable stretch for cutting into a wall and setting new units sits elsewhere in the year. Your account should respect both facts.
Practically, that means pacing against crew capacity rather than against the calendar month. When the schedule is full three weeks out, an aggressive budget just buys you longer lead times and more cancellations. When the board opens up, that is the week to push.
Give the account a monthly conversation about capacity. Which crews are free, what the lead time on ordered units looks like, which product line has room. Then move budget accordingly instead of leaving a flat daily cap running all year.
Avoid the reflex of increasing spend to fix a slow month. Half the time the slow month is a fulfillment or intake problem, and more clicks make it worse.
Cost per booked job, and an attribution window long enough to outlast the committee
Reporting to cost per lead in this market will systematically mislead you, because the approval step sits between the lead and the revenue.
Homeowner associations are common across the valley, and for exterior work an architectural approval frequently sits between the quote and the job. That stretches the time from first click to signed contract well past the default reporting window in most ad platforms.
So instrument the whole path. Click to inquiry, inquiry to booked measure, measure to quote, quote to approval, approval to signed job. Send offline conversion data back into the platform where you can, so the bidding optimizes toward jobs rather than form fills.
Report on cost per booked job and cost per signed contract, segmented by product line and geography. Cost per lead can stay in the deck as a diagnostic, but it should never be the headline.
We do not promise a cost per job before we have your data. Anyone quoting you one has invented it. What we commit to is measuring it honestly and telling you when a segment is not paying, including when the honest answer is to spend less.
Questions we actually get
- What should we budget to test paid search in Las Vegas?
- We will not quote a figure before seeing your product mix, service area and close rate, because a number pulled from another market is fiction. The useful way to size it is backward from the job. Take your average contract value and close rate, decide what a booked job is worth to you, and fund enough clicks in one product line and one geography to learn something. Testing everything at once on a thin budget teaches you nothing.
- Should we run Local Services Ads as well as search ads?
- They are worth evaluating, and they compete for the same phone. If you run both, track them separately and watch for the same inquiry arriving twice. The screening and review requirements are their own project, and eligibility and terms are worth confirming with the platform directly since they change by category and region.
- Is paid social useful for window replacement here?
- It reaches people who have not searched yet, which matters for a purchase homeowners put off for several summers. Treat it as a different job from search: it sells the decision to start rather than catching someone already shopping. Expect a longer path to a booked measure and measure it accordingly.
- How do we stop paying for auto glass and screen repair clicks?
- A negative list built before launch, then a weekly read of actual search terms in the early months. Match types matter less than they used to, so the search terms report is the real control. In this metro we would also exclude commercial and storefront glazing language early, since the resort corridor generates a lot of it.
- Why measure cost per booked job instead of cost per lead?
- Because the gap between them is where the money is. Two campaigns can produce identical lead costs while one fills the install calendar and the other fills the voicemail. With an architectural approval step often sitting between the quote and the job, the lag is long enough that a lead-only view will point you at exactly the wrong campaign.
