Ads for a local service business can bring in their first leads within a week or two of going live, but it takes about 90 days to get a cost per booked job you can plan around. The first month is the expensive one, because the platform is still working out who your buyers are and I'm still cutting the ads that waste money. How fast you get from there to a reliable number depends on your budget, how long your jobs take to close, how quickly you answer new leads, and what the ads are made of.
What happens in the first week of ads?
Very little you can judge yet, because the platform spends the first week testing. Week one of ads also isn't week one of working together. The build comes first. That means the ad account and tracking, the page the ads point to, the follow-up that texts every new lead back within minutes, and the first round of ads. It takes most of the first month, longer when I'm shooting footage at your jobs. Once the ads go live, Meta or Google spends the first days showing different ads to different people to see who responds. Meta's help pages say an ad set needs about 50 results in a week to finish its learning phase, and the ads training I work from uses a number closer to 25. A local business pulling in a handful of leads a week won't reach either one in week one. The first leads can show up within days, and what they cost tells you almost nothing yet.
What does the first month of ads look like?
The first month is usually the most expensive one you'll have, because you're paying for the platform to learn and for me to find out which ads book work. I launch a batch of ads on different angles at once, usually 4 to 12 of them, expecting one or two to win. The losers tell the platform what to stop showing, and you still pay for them. I don't judge a batch before it's run 7 to 14 days, and at local budgets I lean toward the full 14. On Google, a big part of month one goes to reading what people typed in and blocking the searches that were never going to hire anyone, because an account without those blocks can burn 15 to 30 percent of its budget on them. Some first months book a few good jobs and some book close to nothing.
Why does it take about 90 days to get a reliable cost per booked job?
Because the number that pays you is a booked job, and it takes two to three months of leads before there are enough booked jobs to average. Cost per lead usually settles first, often around week three of the ads running and sometimes later. Leads that never book don't pay for anything, though, so a low cost per lead can sit right next to jobs that cost too much. To know what a job costs, enough leads have to get through the quote, the callback and the decision, and bigger jobs close slowly. A replacement or a renovation can sit two to six weeks between the first click and the deposit. Once booked jobs get reported back to the platform, it starts looking for more people like the ones who booked, and that needs a couple of months of data too. My first term with a new client is 90 days for that reason. If you've never pinned down what a booked job costs you, start there.
What should month three look like if the ads are working?
By month three, a working account has a cost per booked job that stays inside a range from week to week, and you know which ads produce it. The losing ads are gone and the budget sits behind the few that book work. New ads go in before those wear out, because any ad gets more expensive once the same people have seen it a few times (I went through that in why your ads stopped working). The weekly numbers stop jumping around. Month three is also when the question moves from whether it works to how much more your crew can take on. If there's room, the budget goes up in steps. Expect the cost per job to climb a little as it does, because the cheapest jobs in your area get bought first, and the extra jobs should still come in under the most you can pay for one and make money.
What makes ads take longer to work?
A budget under the floor slows everything down, because the platform learns from results and a small budget produces too few of them. I recommend no less than $50 a day on ad spend, which is $1,500 a month, and more on Google, where clicks cost more. Spending above the floor gets an account out of the learning phase faster, since the results pile up sooner. Then there's timing. Start ads the same week your busy season opens and you pay for the learning during your best weeks, so I'd start six to eight weeks ahead (booking the slow season has the rest of that). Long sales cycles stretch things too, since a job that takes a month to close takes a month to show up in the numbers. And a brand-new ad account starts from zero. An older account you own keeps the data from everything you've spent before, so the platform isn't starting cold.
Can you make ads work faster?
Yes, a bit, and most of what speeds it up happens on your side of the business. Answer new leads in minutes, because a lead who waits half a day books with whoever answered first, and the ads take the blame for something that happened after the click (speed to lead explains why). Send footage. Ads built from real photos and phone video of your jobs and your crew give the platform more to work with than stock images, and a clip from a jobsite every week or two means there's always a new ad ready before the old ones wear out. Show up to the appointments the ads book, and tell me which leads turned into work, so the platform can look for more people like the ones who paid you. When a client's timeline runs long, the cause is usually one of these, or the page the ads point to, more often than anything inside the ad platform.
How do you know if ads are failing or just still learning?
Look at where the leads stop. If the ads have spent at the floor for a few weeks and brought in almost no leads, the ads or the offer are wrong, and new angles go in. If leads come in and nobody books, the problem usually sits after the ad, in how fast they heard back, the page, the price or the quote. If leads book but each job costs more than it's worth to you, that's what months two and three are for, cutting what doesn't book and moving the money to what does. Each of those needs a different fix, and a report full of clicks and reach can't tell you which one you've got. Be wary of anyone who promises a lead count before they've seen a week of your data. I won't give you one either.
What I put in writing before the ads go live
Before the first invoice, every new client gets the slow version in writing, covering what month one looks like, the fact that it might book very little, and a conservative range for the first 90 days. At day 14 they get a range in writing again, set low on purpose. The first term is 90 days because that's how long it takes to reach a cost per booked job worth making decisions on. If you're thinking about ads for your business and want to see whether it's a fit, book a call.
