How Much Should a Small Business Spend on Google Ads?
Anyone who answers this with a dollar figure before asking about your business is guessing. The real answer is a method: work backwards from the number of leads or orders you need, use published cost benchmarks to set a floor, and let your own results replace the benchmarks within two months.
The question arrives in two forms. "What should I spend?" from an owner who has never run ads, and "Am I spending the right amount?" from one who has been paying an agency for a year and is not sure what it bought. The method below answers both, and it starts with a number that is not a dollar figure.
Start with how many you need, not how much you have
A budget set from what is affordable produces an account that is judged by whether it felt worth it. A budget set from what is needed produces an account that can be judged by arithmetic. So begin there: how many new leads, calls, or orders a month would change the business? Not a hope, a number. A plumber who can take on twelve more jobs a month wants twelve leads that turn into jobs, which at a typical close rate might mean thirty leads. A store that needs forty more orders a month wants forty orders.
Write that number down. Everything else is multiplication.
Multiply by what a lead costs in your industry
WordStream's 2025 Google Ads benchmarks, drawn from 16,446 US search campaigns, put the average cost per lead across all industries at $70.11, up 5 percent on the year, and the average cost per click at $5.26. The spread by industry is the useful part. Those are WordStream's figures, not mine, and they describe averages across thousands of accounts, so yours will differ. As a first estimate they are far better than a guess.
| Industry | Average cost per lead |
|---|---|
| Automotive repair | $28.50 |
| Restaurants | $30.27 |
| Shopping and gifts | $47.94 |
| Beauty and personal care | $60.34 |
| Dentists | $83.93 |
| Home improvement | $90.92 |
| Real estate | $100.48 |
| Apparel | $101.49 |
| Business services | $103.54 |
| Legal services | $131.63 |
The arithmetic: leads needed multiplied by cost per lead for your industry. The plumber who wants thirty leads at the home improvement figure of $90.92 is looking at roughly $2,700 a month. A dental practice that wants twenty new patient inquiries at $83.93 is looking at about $1,700. A law firm that wants fifteen consultations at $131.63 is near $2,000.
For an online store, swap cost per lead for cost per order. If you do not know yours, the checkout leak calculator works it out from your visitors, conversion rate, and ad spend in about a minute.
Check the floor: can the account be measured at that budget?
This is the step most budget advice skips. An ad account is a measurement instrument before it is a sales channel, and it needs enough data to read. My working floor is about a hundred clicks a month, because at typical conversion rates that yields a handful of conversions, which is the least you can draw any conclusion from. At WordStream's average click cost of $5.26 that is roughly $530 a month; in industries with expensive clicks it is more.
If the budget from the previous step is below that floor, do not run ads yet. The account will produce a few clicks, no readable pattern, and a conclusion that "ads do not work for us" that was never tested. That money is better spent on the things that make a smaller budget work later: the Google Business Profile, the landing page, and conversion tracking.
An ad account is a measurement instrument before it is a sales channel, and it needs enough data to read.
Add the learning premium for the first two months
A new account pays for its own education. The first four to eight weeks are spent finding out which keywords bring buyers and which bring browsers, which ad wording earns the click, and which landing page turns the click into a lead. During that period cost per lead runs high, and then falls as the waste is cut. Budget for it: plan on the first two months costing 20 to 30 percent more per lead than your steady state, and do not judge the channel until that period is over and tracking has been working throughout.
Then let your own numbers replace the benchmark
Once the account has collected a few dozen conversions, the benchmark has done its job. Your own cost per lead is now the number, and the budget question turns into a much simpler one: at this cost per lead, and at the rate leads turn into paying customers, how many more do you want to buy this month? If a lead costs $80 and one in four becomes a $1,200 job, every $320 of ad spend produces a job worth $1,200, and the only ceiling on the budget is how many jobs you can take. If the arithmetic does not work at your close rate, no budget fixes it, and the problem is the landing page, the offer, or the sales follow-up, not the ads.
A worked example
A South Florida pool service company wants ten new recurring customers a month. Historically one in three inquiries becomes a customer, so that is thirty inquiries. Home services in WordStream's data run about $91 per lead, so the first estimate is $2,700 a month. Checking the floor: at roughly $5 a click that budget buys around 500 clicks, comfortably above a hundred, so the account can be measured. Adding a learning premium, the first two months are planned at about $3,300, with the expectation that cost per lead falls once the wrong keywords are cut. After eight weeks the account shows a real cost per lead of $64 and a close rate of one in three. From that point the budget is a decision: at $192 per new customer, how many can the crews handle?
Every number in that example except WordStream's is invented for illustration. Your inputs will differ, and the method is the point.
Three ways this goes wrong
The budget is set by what an agency proposes rather than by what the business needs, and nobody can say afterwards what it should have bought. The account is judged before tracking works, so the leads it produced were never counted. Or the budget is split evenly between Google and Meta because that felt balanced, when the two channels do different jobs, convert differently, and deserve separate budgets with separate targets.
If you already run ads and are not sure what the spend is buying, the free 12-point audit has the two checks I would run first: sort campaigns by spend and read the bottom half, and confirm the platform is tracking purchases or leads rather than clicks. If you would rather have someone else read the account, that is what the free consultation is for.
Asked Along With This One
Is there a minimum budget where Google Ads is not worth running?
There is a minimum where you cannot learn anything. If the budget buys fewer than about a hundred clicks a month in your industry, you will not see enough conversions to tell good from bad, and every change is a guess. Below that line, the money is usually better spent on the site or on the Google Business Profile.
Should I split the budget between Google and Meta?
Not by default, and not evenly. Search captures people already looking; Meta interrupts people who were not. They do different jobs and convert differently. Start with search if you sell something people search for, and add Meta once search is measured and profitable, with its own budget and its own target.
How long before I can judge whether the budget is working?
Long enough to collect a few dozen conversions, which at small budgets is usually four to eight weeks, and only if conversion tracking was working from day one. Judging on two weeks of clicks and no conversion data is how most small accounts get switched off right before they would have paid.
What if my cost per lead is far above the benchmark?
First confirm tracking is counting every lead, including phone calls. Then look at the landing page before the bids: a page that does not match the ad wastes clicks at any bid. Only after those two do keyword and bid changes make sense. The benchmark is a yardstick for your industry, not a target for your account.
Not sure what your current spend is buying?
Free 45-minute call. Bring the account and last month's lead count, and I will tell you whether the budget is the problem or something upstream of it is.