How Much Should You Spend on Google Ads? Sizing a Budget From the Math, Not a Guess

Ask ten people how much to spend on Google Ads and you’ll get ten confident numbers, most of them round. $500 a month. $1,500. “Whatever you can afford to lose.” None of those answers is wrong exactly, but none of them is useful either, because a budget isn’t a quantity you pick in the abstract. It’s the output of a small piece of arithmetic that depends on what a click costs in your market, how often clicks turn into customers, and how many conversions the system needs to see before it can do anything intelligent.

Skip that arithmetic and one of two things tends to happen. You set the budget too low, the campaign limps along on a handful of clicks a day, and after six weeks you conclude that “Google Ads doesn’t work for us.” Or you set it too high, early, before your tracking is clean, and burn through a chunk of money feeding a bidding algorithm bad data. Both outcomes are common, and both are avoidable with about twenty minutes of math and an honest look at your numbers.

This post walks through how to size a Google Ads budget from the bottom up. It also covers how a daily budget actually behaves once it’s live (it’s looser than most people think), how budget interacts with Smart Bidding, how to split money across campaigns, and how to change a budget without knocking the account over. Where I cite figures, they come from published sources I’ve linked, and where a number is only an illustration, I say so.

Why “what’s the average budget?” is the wrong question

Search for this topic and you’ll find plenty of agency guides. They tend to land in the same neighbourhood: most small businesses spend somewhere between about $1,000 and $5,000 a month, a single narrow test can run on around $500, and expensive categories like legal need $2,500 or more a month just to collect enough data to optimise. The small business budget guide from Claire Jarrett and the HawkSEM breakdown are two examples of that general shape.

Those ranges are fine as a sanity check. As a decision tool they fall apart, because they average across businesses that have almost nothing in common. A restaurant bidding on a couple of dollars a click and a personal injury firm bidding on nearly ten are both “small businesses.” A shop that converts one in six visitors and a B2B software company that converts one in fifty are both running “lead generation.” Averaging them tells you about the average, not about you.

There’s a second problem with advice that starts from a number. It quietly treats budget as the thing you decide first and everything else as something that follows. In practice it runs the other way round. You start with what a customer is worth to you, work out what you can afford to pay to get one, and the budget is whatever it takes to buy enough of those customers for the account to learn and for you to judge the results. The money is the last number you set, not the first.

The arithmetic that actually sets a minimum

Four inputs get you most of the way there.

• Cost per click (CPC). What you’ll typically pay when someone clicks your ad. Your own account history is best. If you have none, use an industry benchmark as a rough placeholder and expect to be wrong by a wide margin in either direction.
• Conversion rate. The share of clicks that turn into whatever you’ve defined as a conversion: a lead form, a call, a purchase.
• Cost per conversion. CPC divided by conversion rate. This falls out of the first two, and it’s the number that matters most.
• Conversions you need. How many conversions per month you need for the budget to be worth anything, which depends on what you’re trying to do.

Take the first two and divide. If clicks cost $5 and 5% of them convert, each conversion costs you $100. If clicks cost $2 and 10% convert, each conversion costs $20. That’s a five-fold difference in what you need to spend to get the same number of results, and it comes entirely from facts about your market and your website. No budgeting guide can know those facts for you.

Now the fourth input, which is where people most often under-think. How many conversions do you need? It depends on the job the budget has to do.

If you only want customers, the answer is “as many as I can afford,” and the budget is limited by profit. But in the first few months you usually have a different job: you need enough data to learn what works. That’s a separate question from “what can I afford,” and it generally calls for more volume than people expect. Our post on how much conversion data Smart Bidding actually needs goes into the thresholds in detail, but the short version is that the commonly cited guidance for Target CPA is something like 30 conversions in the previous 30 days. Treat that as a rough floor rather than a rule written in stone. Sources disagree about exactly how much is enough, and Google’s own recommendations vary by strategy and over time.

A worked example, with illustrative numbers

WordStream’s 2026 benchmark report puts the average cost per click on Google Search across all industries at about $5.42 and the average conversion rate at about 8.18%. (Those figures are reproduced across several roundups, and the original is the WordStream 2026 Google Ads benchmarks page.) Let’s use them purely as a demonstration of the math, not as a prediction for your account.

• Cost per conversion: $5.42 divided by 0.0818 is roughly $66.
• To get 30 conversions in a month at that cost: 30 times $66 is about $1,990, call it $2,000.
• Per day, that’s about $65, which buys around 12 clicks.

So an account that performs exactly at the cross-industry average needs somewhere near $2,000 a month to produce 30 conversions. That lines up uncomfortably well with the “$1,000 to $2,500 to start” range you see in agency guides, which is probably not a coincidence. But look at what it means in practice: about one conversion a day, on average, from about a dozen clicks. That’s a thin diet for a machine learning system, and it’s lumpy. Some days will have zero conversions and some will have four.

Now change the inputs. Suppose you’re in an expensive category. WordStream lists attorneys and legal services at an average CPC around $9.87, home and home improvement around $8.33, and dentists around $8.00. If a legal advertiser converted clicks at, say, 5% (a number I’m picking for illustration, not quoting from any source), each conversion would cost around $197, and 30 conversions would take nearly $6,000 a month. At the other end, restaurants and food at roughly $2.05 a click, converting at the cross-industry average, would pay about $25 a conversion and need only about $750 for the same 30.

Same target, same method, an eight-fold spread in budget. That’s why a single recommended number is close to useless.

The part that should make you cautious

Benchmarks are averages of averages, and they’re especially unreliable for conversion rate, because “conversion” means different things in different accounts. An account that counts a phone call lasting three seconds as a conversion will show a lovely conversion rate and a terrible business. If your own tracking is shaky, the math above produces confident nonsense. Before you size a budget, it’s worth confirming that your conversions mean what you think they mean. Our guide to auditing conversion tracking is a reasonable place to start, and if you want a sense of what ordinary numbers look like in the first place, see what a “good” CTR, CPC and conversion rate look like in 2026.

Start from what a customer is worth, then work backwards

The calculation above tells you what it costs to buy conversions at market rates. It doesn’t tell you whether you should. For that you need the other side of the ledger: what a conversion is worth.

For an e-commerce store this is relatively direct. Revenue per order, minus cost of goods, shipping, and fees, gives you contribution margin per sale, and your break-even cost per acquisition is that number (or less, if you want some profit). If your contribution margin per order is $40, you can’t sustainably pay $66 to acquire an order from a first-time buyer unless repeat purchases make up the gap. That’s not a Google Ads problem, it’s a unit-economics problem, and a bigger budget will only make it more expensive.

For lead generation the chain is longer. A conversion is a lead, but you get paid for customers. So you need two more numbers: the fraction of leads that become customers, and what a customer is worth. Say, hypothetically, one in five leads closes and the average job is worth $1,500 in gross profit. A lead is then worth about $300 to you, and paying $66 a lead looks comfortable. Change the close rate to one in twenty and a lead is worth $75, and the same $66 is nearly the entire margin. The campaign didn’t change. Your sales process did.

This is also where spam and junk leads quietly wreck the numbers. If a third of your form fills are junk, your real cost per qualified lead is 50% higher than the dashboard says. We wrote about where junk leads come from and how to stop paying for them if that sounds familiar.

Work through this once, on paper, before you touch the budget field. You end up with two numbers. The first is the most you can pay for a conversion and still be glad you did. The second is the cost per conversion you can realistically expect. If the second is higher than the first, no budget size fixes it. You need a better offer, a better landing page, or a different channel. If the second is comfortably below the first, then the budget question becomes a question of how fast you want to go.

How a daily budget really behaves

Here’s something that surprises people who’ve only ever thought about budgets as hard caps. The daily budget you enter in Google Ads is an average, not a ceiling for any single day.

Google’s documentation on managing your spend describes two limits. On any given day, your spend may reach up to twice your average daily budget. Over a month, though, you won’t be charged more than your monthly charging limit, which is your average daily budget multiplied by 30.4, the average number of days in a month. If a campaign goes over that monthly limit through overdelivery, Google applies a credit rather than charging you the difference.

Put numbers on it. A $50 daily budget means any single day can run up to $100, and the month is capped at $1,520. You can see $100 days. You can see $20 days. Over the month, you shouldn’t pay more than $1,520.

Two practical consequences follow.

First, if you only look at one day’s spend, you’ll misread what’s happening. A day at 180% of budget isn’t a bug, and a day at 40% isn’t necessarily a problem. The thing to watch is the monthly total against 30.4 times the daily figure. We covered the diagnostic side of this in why a campaign isn’t spending its budget.

Second, if a bill that’s twice your daily budget on a single day would genuinely hurt, then your budget is too high. Google’s own guidance is to review your settings so you’re comfortable spending up to two times the average daily budget on any day. It’s a useful gut check: set the daily number such that a double day wouldn’t make you wince.

Ad scheduling changed how this works

If you restrict your ads to certain days or hours, the monthly cap logic changed in 2026. Reports from March 2026 describe Google rolling out a change in which campaigns using ad scheduling now pace toward the full 30.4 times monthly limit regardless of how many days the schedule allows. In plain terms, if you only run Monday to Friday, Google may push harder on those days to spend what a full month would have allowed. The old assumption that narrowing your hours would naturally lower your total spend no longer holds. Our write-up on ad scheduling in 2026 covers the details, and the short version for budgeting purposes is this: if you schedule ads for a subset of the week and want a lower monthly spend, lower the budget itself. Don’t rely on the schedule to do it.

Budget and Smart Bidding: the relationship people get wrong

Smart Bidding strategies like Target CPA and Maximize Conversions work by predicting, auction by auction, how likely a click is to convert and bidding accordingly. Budget matters to them in a way it doesn’t matter to manual bidding, because the algorithm can only experiment with money it has.

If your budget is tiny relative to your target cost per conversion, the system has very few chances to find out which auctions convert. Imagine a $20 daily budget with a $60 target CPA. On a normal day the campaign can afford a third of a conversion. Whether it lands one or none is mostly luck, and the algorithm is trying to learn patterns from that noise.

How much headroom is enough? Advice varies. A frequently repeated figure is that Google suggests a daily budget of at least two times your target CPA, with many practitioners recommending three to five times so the campaign doesn’t run dry early in the day. More conservative voices argue for ten times, and some cite older Google guidance as high as fifteen. I couldn’t verify a single current official number from Google’s own pages for the higher ratios, and they’re often repeated from one blog post to the next, so treat the range as an informed disagreement rather than a rule. The direction is the useful part: the closer your daily budget sits to your target CPA, the less the algorithm can learn.

Run that against the earlier example. With a $66 cost per conversion and a $2,000 monthly budget, your daily budget is about one times your cost per conversion. That’s below even the loosest recommendation. It doesn’t mean the campaign will fail. It means a Target CPA strategy is likely to be erratic at that spend, and you should be cautious about reading too much into any given week.

If you’re deciding between strategies, our comparison of Target CPA, Target ROAS and Maximize Conversions walks through when each is a reasonable choice. For low-volume accounts, Maximize Conversions without a target is usually the less brittle way to start, since there’s no CPA number for the system to fail to hit. That’s a judgment call rather than a law, but it fits what the budget math implies.

When your budget is smaller than the math says

Most small advertisers will run the numbers above and discover that the amount they planned to spend is below the amount that would give them a decent shot. That’s uncomfortable but it’s useful information, and there are better responses than quietly proceeding and hoping.

• Narrow the scope until the budget fits. Money spread across ten services and four cities buys almost nothing in each. The same money aimed at your best-margin service in your closest, densest area buys something you can learn from. This is the single most effective lever for a small budget.
• Choose higher-intent traffic. Searches that include a clear purchase or hire signal (“emergency plumber near me”) typically convert better than research-style searches. They’re also often pricier per click, so check that the extra conversion rate pays for itself. Our post on keyword match types in 2026 covers how much control you still have over which searches you appear for.
• Cut waste before adding spend. With a small budget, every irrelevant click is a larger fraction of the total. A tight negative keyword list and a habit of reading the search terms report do more at $1,000 a month than they do at $20,000. See how to build a negative keyword list that cuts waste and how to read a search term report.
• Tighten the location setting. If you serve a defined area, “presence or interest” targeting can show your ads to people who merely searched about your area. That setting might be costing you money, and it matters more the smaller your budget is.
• Run one campaign, not five. Splitting a $1,500 budget across several campaigns leaves each too thin to learn. One well-built campaign with a clean structure is a better use of limited money than a tidy-looking account that starves every part of itself.
• Be honest about the timeline. If the math says you need $2,000 a month and you have $800, you can run it for longer. Expect a slower read on results, and don’t judge it after two weeks.

It’s also fair to conclude that Google Ads may not be the right channel yet. If your unit economics don’t work at the market’s going CPC and a reasonable conversion rate, spending more won’t change that. Saying so isn’t defeatist; it saves you money.

Splitting a budget across campaigns

Once the total is set, the next question is how to divide it. There’s no formula that works universally, but a few principles hold up.

Protect the proven stuff first. If one campaign reliably brings in customers at an acceptable cost, it should get its full budget before you fund experiments. A common failure pattern is the reverse: an exciting new campaign gets generous money while the dependable one runs capped, and the account’s overall return slides without any obvious cause.

Separate by economics, not just by topic. Two campaigns with different profit margins or different target costs shouldn’t share a pool of money, because the cheaper-to-convert campaign will absorb most of it. Our piece on account structure in 2026 argues that fewer, better-organised campaigns tend to beat sprawling ones, and that applies to budget as much as anything.

Brand traffic deserves its own thought. If you bid on your own name, those clicks are usually cheap and convert well, which flatters the blended numbers. Mixing them with non-brand spend can hide that your cold-traffic campaigns are underperforming. We went through whether and how to do that in should you bid on your own brand name.

What about shared budgets?

Google Ads lets multiple campaigns draw from one shared budget instead of each having its own. The appeal is simplicity: one $50 budget across five campaigns instead of five $10 budgets, and the system can move money to where traffic is strongest.

The downside is the same property seen from the other side. Whichever campaign gets the most clicks can dominate the pool and leave the rest with little, and it’s harder to tell afterwards how much each one spent on purpose versus by accident. Commentators who write about this, including WordStream’s budgeting strategies piece, generally frame shared budgets as a convenience for similar campaigns with similar goals, and not a good fit when campaigns have different priorities or you need clean per-campaign reporting. If you’re thinking about pooling at the bidding level as well, portfolio bid strategies have a similar trade-off.

Changing a budget without wrecking performance

At some point you’ll want to raise the budget, ideally because the numbers justify it. The way you do it matters more than most people expect.

Smart Bidding strategies treat large changes in budget or bids as a reason to recalibrate, and during recalibration performance can wobble. Google says as much in general terms. What it doesn’t publish, as far as I can find, is a precise threshold, so numbers you see online come from practitioners’ experience and not from the platform. One roundup suggests keeping increases to roughly 10 to 20% per adjustment, with a week or two between steps; another says changes above 20% are a documented trigger and above 50% can send a campaign back into a full learning period with several days of volatile cost per conversion. Others suggest 20 to 30%. These disagree with each other at the edges, so don’t treat any of them as gospel.

What they agree on is the shape of the advice. Raise in steps, wait long enough between steps to see the effect, and don’t make other big changes at the same moment. If you raise the budget by 30%, rewrite the ads, switch the bid strategy and change the audience in the same afternoon, then whatever happens next, you won’t know why. Our post on the Smart Bidding learning phase explains what resets it and how to tell whether it’s happening.

A related habit worth building: cutting a budget is also a change. Pausing a campaign for a weekend because “nothing’s happening anyway” and then restarting it can send it through another stretch of instability. If you need to reduce spend, reduce it deliberately and by a modest amount, and give it time.

Scale on evidence, not on a good week

The temptation to raise the budget after a great few days is nearly universal, and it’s often wrong. A good week at a low conversion volume is frequently just variance. Before scaling, check that the improvement holds over a longer window, that cost per conversion is stable and not trending up, and that you’re not losing a lot of impression share to rank, which a bigger budget won’t fix. If you want a cleaner read on whether a change helped, Google Ads experiments exist for exactly that purpose, though they need enough data to say anything at all, which loops back to why budget size matters.

Is budget actually your constraint?

Before adding money, confirm that money is the thing holding you back. Two different problems look similar on the surface and have opposite fixes.

If you’re losing impressions because your budget runs out, then more budget buys more impressions. If you’re losing impressions because your ad rank is too low, which is some combination of bid and quality, then more budget changes almost nothing, and the fix is to improve the ads, the landing page, or the bids. The impression share report separates the two: “lost to budget” versus “lost to rank.” Our walkthrough of how to tell the difference is worth reading before you add a dollar.

The “Limited by budget” status has its own quirks, and it doesn’t always mean what it sounds like. We explain what the label does and doesn’t tell you, and a change coming in August 2026, in Google Ads “Limited by Budget”: what it actually means. The short version: a campaign can be flagged as limited by budget and still be a poor place to add money, and a campaign can be perfectly healthy without the flag. Look at cost per conversion and lost impression share together instead of reacting to the label.

Fixed-window campaigns and total budgets

Daily budgets suit always-on campaigns. They’re awkward for things with a start and an end: a launch, a sale, an event, a seasonal push. For years the workaround was manual: divide the money by the number of days, set that as the daily budget, and then adjust every few days as reality drifted from the plan.

In January 2026, Google announced it was expanding campaign total budgets, which let you set a fixed amount for a defined period, to Search, Performance Max and Standard Shopping campaigns in open beta. The window runs from 3 to 90 days, and you have to set both a start and an end date. Google’s announcement says the system takes into account what’s been spent so far and adjusts spending daily to try to use the full amount by the end date, and you won’t spend more than the total. You can read the details in Google’s own post on campaign total budgets.

This is a good fit for anything where the question is “I have $6,000 for these three weeks” and not “I’m willing to pay $200 a day indefinitely.” The caveat is that, like any automated pacing, it still depends on the campaign being able to find enough good traffic. If your campaign is limited by rank or demand, a total budget doesn’t create clicks that aren’t there. For peak-season planning in particular, our Black Friday and Cyber Monday pacing checklist covers the surrounding preparation.

Where daily monitoring fits in

Everything above assumes you set a budget and then watch what it does. That second part is where most accounts quietly fall apart. A budget that was right in March can be wrong in June because a competitor raised bids, a conversion tag broke, a seasonal dip arrived, or one campaign started eating the pool. Nothing announces any of this. The account just keeps spending at the old rate on what is now a different situation.

That gap is what Growera’s Google Ads manager is built around: it monitors your account continuously and reviews it every day, so budget decisions can rest on current data rather than whatever you remembered from the last time you looked. It doesn’t replace the thinking in this post. You still need to know your numbers and set the limits. The point is that drift gets noticed in days rather than at the end of the month. If you’re curious how that works in practice, the ERA AI page explains what the assistant looks at, and the pricing page lays out what it costs.

And if you’re not sure how often manual checking is enough on its own, see how often you should actually check your Google Ads account.

Putting it together: a start-to-finish example

To make this concrete, here’s a made-up local service business. Every number is invented for illustration, so don’t copy them.

A small HVAC company in one metro area wants leads for system replacements. Gross profit on a replacement averages $2,400. They close about one in five qualified quotes, and about half of the form fills and calls are actually qualified (many are repair questions, price shoppers or wrong numbers). So roughly one in ten leads becomes a customer.

• A lead is worth about $240 in gross profit ($2,400 times 0.10).
• They decide to pay up to about $120 per lead, keeping half the margin.
• Their market’s CPC is around $8 and they expect a 6% conversion rate. Cost per lead is then about $133.

That’s slightly above their $120 ceiling. On these numbers the plan is marginal before a dollar is spent. Rather than raising the budget and hoping, they look for improvements: a tighter service area, negatives for repair-only searches, a landing page that asks qualifying questions. Suppose that gets the expected conversion rate to 8% and the CPC down to $7. Cost per lead becomes about $88, which is inside the ceiling with room to spare.

Now the budget. To get roughly 25 to 30 leads in a month at about $88 each, they need around $2,200 to $2,650. Call it $2,400, or about $79 a day, which sits near 0.9 times the cost per lead. That’s below the comfortable range for a Target CPA strategy. So they pick Maximize Conversions with no target, accept that the first few weeks will be noisy, set a calendar reminder to judge the result after six weeks, and agree up front that they won’t change the budget by more than about 20% at a time.

The plan is modest, and it’s written down. If after eight weeks the cost per lead is above $120, they know what to adjust before they spend more. If it’s below, they know how much room they have to scale. That’s the real value of doing the math first: it turns “did it work?” from a feeling into a comparison.

Mistakes I see most often

• Starting with a budget and a hope. Picking $1,000 because it feels reasonable, with no idea what a conversion will cost, means you can’t tell afterwards whether the result was good.
• Judging too early. Two weeks at low volume tells you very little. Decide beforehand how long, and how many conversions, you’ll wait for.
• Treating the daily figure as a hard cap. It isn’t. Days at up to twice the daily budget are normal, and the monthly limit is what you can count on.
• Funding Smart Bidding too thinly. A daily budget that barely covers one conversion makes a target-based strategy erratic. Either raise the budget, narrow the scope, or use a less demanding strategy.
• Raising the budget to fix a rank problem. If you’re losing impression share to rank, more money won’t fix it.
• Changing too much at once. Big jumps, plus other edits at the same time, destroy your ability to learn what caused anything.
• Trusting dashboard conversions blindly. If your conversion tracking counts junk, your whole budget calculation sits on sand.
• Forgetting that the budget needs revisiting. A good number in one month can be a poor one three months later.

The short version

A Google Ads budget is the result of a calculation, not a preference. Divide your expected cost per click by your expected conversion rate to get a cost per conversion. Multiply by the number of conversions you need, roughly 30 a month as a commonly cited floor for target-based Smart Bidding, to get a sensible monthly minimum. Then check that cost per conversion against what a conversion is actually worth to you. If it doesn’t fit, the answer is a better offer, a tighter scope or a different channel, not more spend.

Once the campaign is live, remember that your daily budget is an average. Daily spend can reach twice that figure, the month is capped at 30.4 times it, and ad scheduling no longer reduces monthly spend on its own. Give Smart Bidding enough room to learn, split money by economics rather than topic, and change budgets in modest steps with time between them. Use total budgets for fixed windows. Before adding money, check whether you’re limited by budget or by rank, because the fixes are opposites.

Finally, write the numbers down and set a date to review them. The accounts that do well over a year aren’t usually the ones that guessed the right number on day one. They’re the ones whose owners knew what they expected, noticed early when reality drifted from it, and adjusted before the month ran out.