The mistake that burns most first-time budgets
The single most expensive mistake in Facebook advertising is boosting posts from your business page and calling it a campaign. Boosted posts have limited targeting, limited optimization options, and no conversion tracking. You are essentially paying Facebook to show your post to a slightly larger audience with no mechanism to measure what, if anything, that exposure produced.
Instead, run your campaigns through Meta Ads Manager. It is free to use, and it gives you access to detailed targeting, proper conversion tracking, and the optimization algorithms that actually make Facebook advertising work. The learning curve takes about an hour. That single hour will save you hundreds of rand compared to boosting posts blindly.
Start with the offer, not the audience
Most businesses jump straight to targeting and creative before defining their offer, and that is backwards. Facebook can find the right people with impressive precision, but if your offer is "call us for a free quote," you are competing with every other business saying the exact same thing.
Strong offers for service businesses are specific and low-commitment. A free 21-point roof inspection, a R49 drain assessment (waived if they book), an annual electrical safety check with a concrete price. These perform because they give someone a sensible reason to respond right now. According to WordStream, the average conversion rate for Facebook ads across all industries is about 9.2%, but service businesses with specific offers consistently outperform that average because the ad gives people a clear, low-risk next step.
Targeting that does not waste spend
For local service businesses, your most important targeting parameter is geography. Set your location to the specific cities or postal codes you serve, not an 80 km radius around your office. A plumber spending ad rand reaching people 40 km away is paying for impressions that will almost never convert into a job.
Layer on demographic filters like property ownership or household income if they are relevant to your service (Meta allows both through standard demographic targeting in many markets). Beyond that, keep initial targeting relatively broad and let Facebook's algorithm find the right people within your geographic area. Overly narrow interest-based targeting limits the algorithm's ability to optimize, and the data from Meta's own case studies shows that broader audiences with proper conversion optimization outperform hyper-targeted campaigns in most cases.
Creative that stops the scroll
Your ad creative has about 1.5 seconds to earn attention in a Facebook feed. That is not a made-up number. Meta's own research on mobile ad attention confirms it. In that window, your image or video needs to communicate something relevant to the viewer.
What works for service businesses: before-and-after photos of real jobs, short videos (under 30 seconds) showing your team at work, images of finished projects with text overlay stating the offer. What does not work: stock photos, corporate graphics with your logo dominating the frame, walls of text that nobody will read at feed-scrolling speed.
Run at least three to four different ad creatives from day one. Facebook will automatically show the better-performing versions more often, and you will learn what resonates with your audience without guessing. Some of the best-performing ads for service businesses are surprisingly simple. A phone photo of a completed job with an honest caption outperforms polished graphics more often than you would expect.
Track what matters and cut what does not
After your first week, look at cost per lead, not impressions or reach or click-through rate. Those vanity metrics feel good but do not tell you whether you are making money. If you are running lead form ads, the cost per lead is visible directly in Ads Manager. If you are sending traffic to a landing page, make sure you have the Meta pixel installed and a conversion event firing when someone submits your form.
Give each ad creative at least 3-5 days and R50-R75 in spend before making judgments. Turning ads off after one day because they "did not work" does not give the algorithm enough data to optimize. Once you have a week of data, pause the creatives that are not producing leads and shift that budget to the ones that are. This iterative process, test, measure, cut, reinvest, is how you turn a R1,000 test budget into a system that reliably generates leads month after month.