Google Ads for Christmas Light Installers: Budget Math

How should the campaign be built?#

One Search campaign, called Seasonal Install, and nothing else in the first season. No Display, no Performance Max, no broad match; Strandr's setup video (September 5, 2026) gives the same advice and it is right. The settings that matter:

Location. Google's default is "Reach people in, regularly in, or who've shown interest in your targeted locations." Switch it to presence only, which Google describes as "Reach people in or regularly in your targeted locations," on the advanced location options page. A homeowner in another province who has "shown interest" in Ottawa is not a lead. Target the same neighbourhoods you listed as service areas on your Google Business Profile.

Schedule. Google's ad scheduling page lets you "Specify certain hours or days of the week when you want your ads to show" and "Set bid adjustments to increase or decrease your bids for specific days and times." Run 7 am to 9 pm. If nobody answers the phone after 6, bid down the evening rather than off, because the form still works.

Conversions. Two: a phone call from the ad or the landing page, and the form submission. Bid for conversions once there are 30 of them; before that, bid for clicks with a cap at the Keyword Planner number.

Ad groups. Four, one per way a homeowner types the search:

The cost group is the one most installers skip because they do not want to print a price. Print it. A searcher typing "cost" is comparing, and the ad that says "from $8 a foot, takedown included" gets the click from the one that says "call for a quote." That $8 is GlowBid's default C9 roofline rate; put your own in.

Which negative keywords stop the bleeding?#

Google's negative keywords page says they "let you exclude search terms from your campaigns," with negative broad as the default, plus negative phrase and negative exact. The catch in Google's own words: negatives do not add synonyms or singular and plural variations for you, so enter both forms. Start with this list as a shared negative list applied to the campaign, and add to it every Monday from the search terms report.

Forty-odd terms, and the retail group does the most work. In the first two weeks, read the search terms report every morning; a single day of "christmas light show tickets" clicks at $8.33 pays for a ladder stand-off.

Where does the click land?#

On one page built for it, not your homepage. What the page needs, top to bottom: the offer in one line (a priced mockup of your house within a day), three dusk photos of your own installs, one mockup on a real customer's photo labelled as a mockup, the starting price per foot, the phone number, and one form that asks for a photo of the front of the house. Then the line that is on every quote anyway: photo measurements are estimates, confirmed on site.

The form is the whole point. A homeowner who sends a photo gets back a drawing of their own house with the lights on it and a price from your rate card, and that is the quote they show their spouse. How to make a Christmas light mockup walks through it, and the free roofline tool does the drawing and the footage with no login until export. Get the mockup back the same day; Geiman's guide puts the close rate at 15 to 25 percent when the quote is out within 2 to 4 hours.

Worked example: a $3,000 season at three conversion rates#

Fifty dollars a day from September 15 to November 13, 60 days, $3,000. Clicks at LocaliQ's $8.33: 360. Everything below the click row depends on your landing page and your close rate, so here it is three ways. The 20 percent close is the middle of Geiman's range; the 4 and 2 percent landing pages are example assumptions of mine, there to show where the line is.

Rounded to whole leads and jobs, and conservative on the jobs row. The first column is a profitable season on the default rate card: 5 jobs, $10,670 of revenue, $3,350 net after the ads, and 5 new houses that re-book next year with no ad spend at all. The third column is the same $3,000 buying one job. The difference is not the bid; it is whether the page shows the searcher their own house with lights on it and asks for one photo.

Strandr's video uses $250 per lead as the example of a target you could give Google, and warns that a target caps volume on a new campaign. At a 20 percent close that $250 is $1,250 per signed job, which on our card is $20 short of the entire profit. Set the cap from your own P&L, not from an example.

When do you turn it off?#

Three rules, checked every Monday:

1. The calendar is full. One crew installs about 2 two-storey houses a day, 80 in an 8-week window on our route planning numbers. When the booked jobs match the days you have, every further lead is one you will quote and turn away.
2. Cost per signed job has been over the cap for 14 days. Two weeks is enough clicks to know it is the page and not the weather. Fix the page first, then the negatives, then the bids; if it is still over, stop.
3. Ten days before the last install date. If your last install is December 15, a lead on December 6 cannot be quoted, signed, supplied and installed in time, and the ad money buys an apology.

Turn it on again in February, at $10 a day, with the permanent ad group only. Permanent lighting searches run all year, the ticket is more than two and a half times the seasonal one on the default card, and a February lead is an April install with no competition for the crew's time.

Mistakes that waste an ad budget#

• Sending the click to the homepage. The searcher wanted a price and a picture. They got a menu.
• Broad match on "christmas lights." It buys the light show, the Amazon strand and the DIY clip searches, and the negative list only catches what you thought of.
• Presence or interest left on the default. Clicks from people who are not in your service area and never will be.
• No negative list on day one. The first week's search terms report is where the retail and show clicks show up; without negatives they keep showing up.
• Starting in November. The searches start in September; by November you are bidding against every installer whose calendar is also full.
• Running past the last install date. A December 10 lead is a quote you cannot fulfil.
• Judging it on cost per click. A $12 click that converts at 8 percent is cheaper than a $6 click at 2 percent. Judge it on cost per signed job.
• No follow-up on the leads it buys. A $103 lead that gets one call and silence is a $103 lead someone else closes. Our 21-day quote follow-up sequence is five touches, tied to the quote's status.

GlowBid does not run ads. What it does is turn the lead the ad bought into the thing the ad promised: a mockup on the homeowner's own photo, footage by run, a price from your rate card, a quote PDF they can sign on their phone and a Stripe deposit, the same day. Core is $299 a year at the founding price, 7-day trial, no card; Pro at $499 adds the pipeline and the automated follow-up that keeps the ad's leads from going cold. Full product open November 1, 2026, 100 founding places.