How to set a Google Ads budget for a local service business

Most budget advice hands you a dollar range and leaves you to guess where in it you belong. A business whose average job is worth $2,000 and a salon whose average ticket is $80 cannot share a number, and neither needs one borrowed from somebody else's account. You already hold the inputs that decide it: what a job is worth to you, how often a lead turns into a job, and how many jobs you can take on this month. This article builds the budget from those numbers, shows where the money goes on Google now that call-only ads and Local Services Ads have both changed, and lists what to read in the account before touching the number again. Google's own rules quoted below were read on Google Ads Help on September 13, 2026, and are dated in the text.

One thing this article does not repeat. How Google paces an average daily budget over a month, whether it publishes a minimum budget, and how many clicks a first Search campaign needs before its numbers mean anything are all written up at aevin.dev/blog/google-ads-for-small-business, together with the account setup that should exist before any budget does. This one starts from the other end, the money: what a lead is worth to your business, and what that makes the monthly figure.

Work out your target cost per lead before your budget

A budget is an output. The input is what one lead is worth to you, and three numbers decide it. Write them down before you open Google Ads.

1. 1.Average job value. Not your best job: the average of what a new customer actually pays you on a first job, or over the first year if customers in your trade come back.
2. 2.Gross margin on that job. What is left after labor, parts, materials or chair time. Advertising is paid for out of margin, not out of revenue.
3. 3.Close rate on leads. Out of ten people who call or fill in the form, how many book and pay. If you have never counted, count for two weeks before you spend anything.

Now the arithmetic. Margin per job times close rate is the margin one lead brings in on average, and it is also the most a lead can cost before the jobs it produces start losing money. Then decide what share of that you are willing to hand to advertising. If you will spend a third of it, your target cost per lead is a third of the margin per lead. That number is yours rather than a benchmark, and it is the one that tells you whether a campaign is working. The full version of this arithmetic, including why cost per lead and cost per acquisition are different numbers, is at aevin.dev/blog/cost-per-lead-vs-cost-per-acquisition.

A worked example in round numbers. An average job of $1,000 at a 50 percent margin leaves $500 per job. At a 20 percent close rate one lead in five becomes a job, so each lead is worth $100 of margin on average, and $100 is the break-even price of a lead. Spend a third of that and the target is about $33. Leads at $30 are inside your limit. Leads at $90 still make money, but only $50 on each job, because the five leads behind it cost $450. At $120 a lead, each job costs $600 to win against $500 of margin, and the campaign loses $100 on every job it books, however good the clickthrough rate looks.

Target cost per lead is a decision, not a market rate. Nobody can tell you in advance what a lead will cost for your service in your area. You can set the ceiling, spend, and compare.

Turn cost per lead into a monthly and a daily budget

Decide how many new jobs you want in a month, work backwards through the close rate to the number of leads, and multiply by your target cost per lead.

1. 1.Jobs wanted per month. Be honest about capacity: two extra installs a week is about eight a month.
2. 2.Leads needed. Jobs divided by close rate. Eight jobs at a 20 percent close rate need 40 leads.
3. 3.Monthly budget. Leads times target cost per lead. Forty leads at $33 is $1,320 a month.
4. 4.Average daily budget. The monthly figure divided by 30.4, which here is about $43 a day.
5. 5.A cash check. If the monthly figure is more than you could carry through two months in which nothing closes, lower the job target and let the budget follow it down, rather than keeping the target and cutting the daily budget below what the plan needs.

The 30.4 is not a rounding habit. Google's spending limits page, read on September 13, 2026, puts the monthly spending limit for most campaigns at 30.4 times the average daily budget, 30.4 being the average number of days in a month. Dividing by it makes the daily figure you type in and the monthly figure you planned describe the same money. How Google paces that money from one day to the next is covered in aevin.dev/blog/google-ads-for-small-business and is not repeated here.

If you would rather start from a revenue goal than from a number of jobs, the ad budget calculator at aevin.dev/tools/ad-budget-calculator runs the same chain from revenue to sales, booked appointments, leads and spend. It adds a booking step between the lead and the sale, and it works from the cost per lead you pay today rather than from a target. Its daily figure assumes a 30 day month, so for a Google daily budget take its monthly figure and divide by 30.4.

One input stays loose on purpose: how much search demand exists for your service where you work. You cannot know it before spending, but you can check it. Google's Keyword Planner page, read on September 13, 2026, says the tool shows estimates of the number of searches a keyword gets each month, and that a plan forecast shows how many conversions, clicks or impressions you are likely to get for your keywords based on your spend, with conversions available once conversion tracking is on. Enter your monthly figure and compare the forecast with the leads you need; a forecast far below your target is worth knowing before launch. The same page names your bid, budget, ad quality and location targeting among the things that shape real results, so read the forecast as a check rather than a promise, and trust two weeks of your own account over it.

Where the money goes: Search, calls and pay-per-lead

For a business with a service area, Google spend comes down to three lines. Two of them changed in 2026, so older advice about them is out of date.

Search campaigns

Text ads shown against what people type when they have the problem now. At the start this is where most of a local service budget belongs, because it is the line where you choose the keywords, pay per click, and can see in the search terms report which phrase produced a lead. The whole calculation above assumes this line.

Phone calls: call assets, not call-only ads

If you planned a separate call-only campaign for the phone, drop that line. Google's help page on call ads, the format built to appear only on devices that make phone calls, read on September 13, 2026, lists February 2026 as the date all options to create new call ads are removed and February 2027 as the date existing call ads stop receiving impressions, and tells advertisers to add their phone numbers to responsive search ads as call assets instead. The budget consequence is simple. Google's call assets page says a tap on the call button, or a click on the "Call us" button on desktop, costs the same as a click on the ad's headline, a standard cost per click. Calls therefore come out of the Search campaign's budget, and they belong inside the Search figure rather than in a line of their own.

Counting those calls takes one setting. Google's page on phone call conversion tracking says you set a minimum call length and every call that lasts at least that long is counted as a conversion, and its call reporting page adds that a phone call itself has no duration requirement, so reports will show more calls than call conversions. Set the minimum to the length a real enquiry takes in your business. A 12 second wrong number should not count as a lead, and it should not teach the bidding to buy more calls like it.

Local Services Ads, now moving into Google Ads

Local Services Ads charge per lead rather than per click, and they are in the middle of a move. Google's transition page, read on September 13, 2026, says existing Local Services Ads campaigns will be automatically migrated to a specialized Performance Max campaign type optimized for pay-per-lead goals, and that you still pay only for valid leads, such as phone calls and messages, rather than ad clicks. It dates the first phase to August 2026 for select home and storefront service advertisers in the United States, naming trades such as plumbing, HVAC, roofing, pest control and moving. Broader groups, including service-area businesses without physical storefronts, follow in late 2026, and non-U.S. accounts and all remaining categories in 2027. Not every business can join. Google's getting started page says Local Services Ads are available only for certain service categories, with an eligibility check for your area, and the transition page says new campaigns with pay-per-lead goals cannot be created inside Google Ads at this time.

Three details on the transition page matter for a budget. Your historical average weekly budget is divided by 7 to become an average daily budget, and monthly spend is capped at that daily average times 30.4, the same multiple Google uses for most campaigns. Manual bidding, such as setting a maximum cost per lead, is no longer supported once the campaign is in Google Ads. And these campaigns are keywordless, working from your Google Business Profile, so there is no keyword list to steer them with. Treat pay-per-lead as its own budget line with its own cost per lead, and compare it with Search on cost per booked job rather than on the price of a lead alone.

A workable starting shape for a $1,300 month is nearly all of it in one tightly targeted Search campaign with call assets on its ads, and pay-per-lead as a separate line only if your category qualifies for it. Splitting $1,300 across several campaigns leaves each of them too thin to learn from.

Check that the budget can feed the bidding

Automated bidding learns from conversions, and it buys those conversions with your budget. Google's page on the learning period, read on September 13, 2026, says it can take up to around 50 conversion events or 3 conversion cycles for a bid strategy to calibrate to a new objective, although it can be faster with more conversion data, and that the learning period is not applicable to Manual CPC. Among the factors that set its length, the page lists the number of conversions and the duration of your conversion cycles, meaning the time it takes for a click to result in a conversion.

Turn that into dollars. Fifty conversions at a $33 target cost per lead is $1,650 of spend. A $1,320 month buys about 40 leads, so it gets there in roughly five and a half weeks, if all 50 are needed. A $600 month buys about 18 leads and takes close to three months. Fifty is Google's upper estimate rather than a quota, but when the budget needs a quarter to reach it you have two honest options: narrow the campaign so the same money buys conversions faster, or start on Manual CPC, where the learning period does not apply, and accept that you are doing the steering. Which edits restart learning once it is under way is covered in aevin.dev/blog/google-ads-for-small-business.

Conversion cycle length matters for the same reason. An emergency call can turn into a booked job within the hour, while a quote for a large job may take weeks to become a signed one. The longer the cycle, the longer before the numbers mean anything, and the more weeks of spend the budget has to carry before you judge it.

Count the right conversion. If every form fill counts, spam included, the bidding learns to find more spam. Count booked calls and leads you would call qualified, and keep everything else out of the conversion column the budget is judged on.

Keep the budget inside your service area

Location is one of the quieter ways a budget leaks. Google's location targeting page, read on September 13, 2026, lists a radius around a location among the areas you can target. For a business with a van and a drive time, a radius around your address usually fits better than a city name, because a city boundary does not match how far you will actually travel. Whether Google should also show your ads to people who are only interested in your area is a separate setting, covered in aevin.dev/blog/google-ads-for-small-business.

• –Set the radius to the distance you will really travel for a job, not the one you would like to. If you would not drive 40 minutes for a $120 job, do not pay for clicks 40 minutes out.
• –Do not shrink it to nothing. Google notes that selecting a small location could mean your ads show only intermittently or not at all, and radius targeting requires a radius of at least 1 km around a location.
• –If the radius takes in a town you do not serve, exclude that town. Google's page on location exclusions says that when an area is both included and excluded, the exclusion takes priority.
• –After two weeks, open the matched locations report, which shows performance by the locations that matched your ads, and see which towns produced leads and which only produced clicks.

What to read in the first weeks, before changing spend

Reading is not editing. The first weeks are for looking at the account often and changing as little as possible, roughly in this order.

1. 1.Whether conversion tracking records anything at all. Clicks and zero conversions after a week, with a phone that rings, point at the tracking before the ads.
2. 2.The search terms report. Not your keyword list but the phrases people actually typed. This is where you find out you are paying for jobs you do not do.
3. 3.Impression share, and why it was lost. Google defines Search lost IS (budget) as the percentage of time your ads were not shown on the Search Network due to insufficient budget, reported at the campaign level only, and Search lost IS (rank) as the share lost to poor Ad Rank in the auction. The first points at the budget. The second points somewhere else.
4. 4.Cost per lead against your target. One number, the one you worked out before you started.
5. 5.Matched locations. Which parts of the radius produced leads.
6. 6.Call details, if calls matter to you. Google's call reporting shows each call's start and end time and whether it was missed or received. A missed call is usually a click you paid for and a lead nobody took.
7. 7.How fast a form lead gets a call back. Google cannot see that, and a slow callback wastes a lead the budget has already paid for.

When to raise the budget, and when the budget is not the problem

Raise the budget when the account says it is running out of room and the leads are inside your target. Do not raise it because a month felt slow.

Two failure modes look alike from the outside and are opposites inside. A budget too small for the bidding to learn shows up as a campaign that spends its budget most days, conversions in single digits and a cost per lead that swings from week to week. A targeting or offer problem shows up as budget the account cannot spend, or clicks with no leads behind them. Adding money to the second only buys more of the same clicks.

Budget will not fix a landing page that does not convert or a phone that rings out at lunchtime. Before spending another dollar, call your own number from a mobile at 12:30 and see what happens. What a page owes the visitor an ad sent is written up at aevin.dev/blog/landing-page-for-facebook-ads, and what to do in the first minutes after a lead arrives is at aevin.dev/blog/how-to-follow-up-on-leads. Both were written about Meta traffic, and the parts about the page and the phone carry over to a search click.

Common mistakes that waste a local service budget

• –Cutting the budget on day three because one day spent more than the daily figure. That is normal pacing inside the monthly limit, explained in aevin.dev/blog/google-ads-for-small-business.
• –Changing bids, budget and keywords in the same week, so no result can be traced back to any one change.
• –Splitting a small budget across several campaigns so that none of them reaches enough conversions to learn from.
• –Counting every form fill and every short call as a lead, and then judging cost per lead on that count.
• –Targeting a whole metro area when the van goes 15 minutes.
• –Planning a call-only campaign now. Google's call ads page lists February 2026 as the date the options to create new call ads are removed, and phone numbers go on Search ads as call assets instead.
• –Borrowing Meta's budget math for Google. The Meta version of this calculation, built on Meta's weekly learning threshold, is at aevin.dev/blog/how-much-budget-for-meta-ads. Google's figure of up to around 50 conversion events is an estimate for calibration rather than a weekly count, so the same arithmetic does not carry over unchanged.
• –Treating the price of a click as the metric. The metric is cost per booked job, and the click price is only one part of it.

Where Aevin fits, and what it will not do

Aevin is our own product, at $97, $197 or $497 a month. It connects Meta, Google Ads and TikTok in one dashboard, with each platform on its own row, and on this subject the precise version is more useful than a pitch. For Google it shows spend, leads and cost per lead beside Meta and TikTok, which gives the target from this article a number to be compared with. The lead count for Google is the conversions figure Google reports, rounded to whole leads, so it is only as honest as the conversion actions your account counts, and the minimum call length matters there too. Impression share and the matched locations report are not in Aevin, so those stay in Google Ads. The daily agent pass proposes changes, and on a running Google campaign those changes are pause, resume, the daily budget and the bid strategy. None of them is applied until a person approves it, shown as the value before and the value after. Aevin builds Search campaigns, created paused, for leads or website traffic. It does not build Performance Max, video or app campaigns, so the pay-per-lead line lives in Google's own interface. And it makes no promise about what your leads will cost.