Most people ask this question backwards. They decide what they are comfortable spending, put that number in, and then judge Google Ads on whether it worked.
The problem is that your comfort level has nothing to do with what your market costs. If a click in your industry costs eight dollars and you need roughly forty clicks to get a customer, then a customer costs you around three hundred and twenty dollars in ad spend before anything else. A budget of two hundred a month does not buy you a slow start in that market. It buys you nothing, spread thin enough that you also cannot tell whether it would have worked.
Work out the number from the other end
You need three figures, and you already have two of them.
What a customer is worth to you. Not the first invoice. The total, including repeat work and referrals, minus what it costs you to deliver. A tint shop where a customer spends four hundred dollars once is a different business from one where the average customer comes back twice and sends a friend.
How many enquiries it takes to close one. Count it. Most owners guess a much better number than the real one, because the enquiries that went nowhere are the ones you forget. If one in four people who call become customers, that is your number.
What a click costs in your market. This is the one you have to look up. Google’s Keyword Planner will give you a range for your services in your area, free, before you spend anything.
From those three: multiply the clicks needed per enquiry by your close rate, and you get roughly what an acquired customer costs in ad spend. Compare that to what a customer is worth. If the ad cost is more than the customer is worth, the answer is not “spend less” - the answer is that this channel does not work for you at these numbers, and you should find that out on a spreadsheet rather than over six months.
The floor nobody tells you about
Below a certain spend, ads do not fail, they just fail to tell you anything.
You need enough clicks to know whether the ones you are buying are the right ones. A campaign that gets eleven clicks in a month has not been tested. It has been sampled, badly, and any conclusion you draw from it is noise.
As a rough floor, work out what forty to fifty clicks a month costs in your market and treat that as the minimum for learning anything. For a lot of local service businesses that lands somewhere in the several hundreds per month. If that is more than you want to risk, that is a completely reasonable decision - but the right response is to not run ads yet, not to run them at a level that cannot teach you anything.
Where the money actually goes wrong
In our experience the budget is rarely the thing that breaks a campaign.
Sending everyone to the homepage. Someone searching for one specific service lands on a page describing seven, and has to go looking. A meaningful share of them do not. You paid for that click either way.
No call tracking. If the phone rings and nobody logs where it came from, you cannot tell profitable campaigns from expensive ones. You end up judging the whole account on gut feeling, which usually means turning off the thing that was working because it looked expensive.
Broad match with no negative keywords. You end up paying for searches from people looking for jobs, for DIY instructions, or for free versions of what you sell. This is the single most common way a small budget gets eaten before it reaches a buyer.
Running ads to fix a website problem. If the site does not explain what you do quickly, ads do not solve that. They just buy more people the chance to leave. This one is worth being blunt about: if your conversion rate is the problem, more traffic is the most expensive possible fix.
A sane way to start
Pick the one service with the clearest buying intent, not the one with the most searches. Someone typing “emergency ac repair near me” is closer to paying you than someone typing “how does ac work.”
Send those clicks to a page about that specific service. Set up call tracking before the first click, not after the first confusing month. Add negative keywords in the first week and again in the second, because that is when the junk shows up.
Give it enough budget to produce real numbers, and enough time to produce enough of them. Then decide with the data instead of the feeling.
When ads are the wrong first move
If people cannot find you when they search your own company name, if your Google Business Profile is half filled in, or if enquiries currently sit unanswered for a day, ads will amplify all of it. Paid traffic is a multiplier. It multiplies whatever is already happening, including the bad parts.
Fixing the free stuff first is boring advice and it is usually correct.