Facebook Ads for Michigan Service Businesses That Actually Work

Despite all the noise about iOS privacy, Apple Mail Privacy, and ad-blocker adoption, Facebook ads still produce real, measurable jobs for Michigan service businesses in 2026. They work less well than they did in 2018. They work much better than the typical owner who tried them once and gave up assumes.

The difference between a campaign that produces 30 dollars cost-per-lead and one that produces 250 is almost never the creative. It is the setup, the targeting, and the offer.

For a Michigan service business with under 5,000 dollars a month in Facebook budget, lead form ads (the kind where the form opens inside Facebook, not on your website) outperform website conversion campaigns by 30 to 50 percent on cost-per-lead. The reason is mechanical: the form pre-fills with the user's name, phone, and email from Facebook, which removes the friction that kills mobile form completion.

The trade-off is lead quality. Lead form leads are slightly lower intent than website form leads. Plan to call them within 5 minutes. The companion piece on tracking phone leads covers the response side.

For most Michigan service businesses, the targeting that produces real results is simple: a tight geographic radius (10 to 25 miles from the shop), a homeowner audience (Facebook still has a 'homeowner' interest signal that works), and the relevant age range for the service. That is it. No third-party data, no lookalike audiences for the first 60 days.

After 90 days of real lead data, build a 1 percent lookalike audience from the leads who actually became customers (not all leads). This is the single highest-impact targeting move available.

The highest-performing Facebook ads for Michigan service businesses are short videos (15 to 30 seconds) shot on a phone, showing the actual technician on an actual job in Michigan. Local landmarks in the background help. Generic stock footage destroys performance.

Static image ads still work as a complement. The format that performs: real photo of a real completed job (before and after), one line of copy ('We replaced a 14-year-old water heater in Royal Oak this morning'), and a clear CTA.

Below 1,500 dollars a month, run one campaign with one ad set and three creatives. Do not split-test 10 campaigns on a small budget. The data is too thin and the algorithm cannot optimize. One campaign with steady spend produces better results than scattered tests.

Above 3,000 dollars a month, you can run a second campaign with a different offer (a maintenance plan, a financing promo, an emergency service). Above 5,000, retargeting and lookalikes become worth the additional complexity.

Stop running boosted posts from the page. The boosted post is the worst form of Facebook advertising. It targets the wrong people and reports the wrong metrics. Always run real campaigns from Ads Manager.

Stop running ads to your homepage with no offer. The homepage is for branded search visitors who already know you. Paid traffic needs a specific landing page or a lead form with a specific offer.

Stop running ads without a tracking pixel installed. The pixel is free, takes 20 minutes to install, and changes everything about what the algorithm can do for you.

Most Michigan owners pull Facebook ad campaigns at day 7 because the numbers look bad. Day 7 numbers are always misleading. Facebook needs 50 to 100 conversion events to stabilize an audience. For a service business at 30 leads a month, that is 8 to 12 weeks of steady spend, not 1.

Set a 90-day budget, run the campaign without major changes for 90 days, and decide at day 91 whether it worked. The shortcut is to fail.

MBI does not run paid ads as part of the standard program. The MBI site and GBP work make every paid ad campaign perform better, because the landing page, trust signals, and review profile carry the click forward. Review the program page, the pricing, and the HVAC web design guide if you want the trade-specific conversion details.