bro you have already gotten the budget numbers from rest of the guys,
so i wont repeat them lol, but i think everyone here including you is answering the wrong question. you asked how much to spend. the question that actually decides this is how many jobs can you physically do in a week by yourself, and whats the average ticket on one.
work it backwards. say you can comfortably do 4 or 5 jobs a day solo, so call it 25 a week, and realistically you want maybe 15 to start while you're still doing quotes and invoicing and everything else yourself. at surfsideinbound's realistic number of $125-150 a lead and lets say you book half of them, 15 jobs a week costs you somewhere around $4,000 a month in ad spend.
Now the only question left is whether 15 garage door jobs a week clears $4,000 plus your own pay plus parts. if your mix is mostly spring and opener repairs at two or three hundred bucks, thats extremely tight. if theres a decent share of full door replacements at a couple thousand, its comfortable. you know your ticket mix and nobody in this thread does, so nobody in this thread can actually tell you if $50 or $250 is right.
which brings me to the thing i'd genuinely be more worried about than budget.
You said that you handle the calls yourself.
Thats the part that will quietly eat this whole experiment. a garage door search is an emergency search, the door is stuck and the car is trapped and the person is late for work. they are calling three companies in a row and hiring whoever picks up. so if you're up a ladder with a drill in your hand when the phone rings, that click cost you $30 and produced nothing.
At $250 a day you'd be buying roughly 8 clicks a day and you'll be on site for most of them.
thats not a reason not to run ads, its a reason to sort out call answering before you turn them on. cheapest fix is an answering service that takes the job details and texts them to you, they run maybe a couple hundred a month for low volume and they pay for themselves the first time they catch a call you'd have missed. After that, a text-back-in-30-seconds on any missed call. and dont bother with a fancy CRM, a notebook is fine, but you do need to know which calls came from ads and how many turned into money, otherwise in two months you'll have spent $6,000 and have a feeling instead of a number.
on channel, the people saying start with LSA are right and id go further, id put the whole budget there for the first month and not run search at all. local services ads charge you per lead instead of per click, so you are not paying $30 to find out someone was shopping for a part on amazon, and you can dispute a junk lead and get the money back. also the google guaranteed badge sits above everything else on the page,
and for an unknown one man business that badge is doing more work than any ad copy you could write. get the background check and licence stuff done, get five reviews from people whose doors you've already fixed, then look at whether search is worth adding on top.
last thing, harambetoosoon's point about the radius is the best thing in this thread and its being underrated. 20 miles around atlanta is a million households and you cannot serve a million households, you're one guy in a van. every mile of that radius you dont want is money spent showing your ad to someone whose job you'd have to drive 45 minutes to reach. id pick maybe 8 to 12 zip codes,
ones with detached houses and decent home values and a drive you'd actually be happy making, and start there. tighter area on a small budget also means you show up more often to the same neighbourhood instead of once to everybody, which matters more than people think.
