How much should you spend on Google Ads? Work out your floor first

Almost every answer to this question is a range borrowed from what other businesses spend. That tells you nothing about whether your budget can work, which is the only thing you actually need to know before committing. There is a way to calculate it, and it takes about two minutes.

The usual answer is useless, and here is why

Search "how much should I spend on Google Ads" and you get ranges: $1,000 to $2,500 a month, or $1,000 to $10,000, or "it depends on your industry". None of those is wrong, and none of them helps, because a range derived from what other businesses happen to spend tells you nothing about whether your budget can work.

There is a better question. Google Ads has a documented mechanical requirement before its bidding can do its job. That requirement, combined with two numbers specific to you, produces a floor — the point below which spending more carefully does not help, because the system cannot learn. Budget above the floor is a strategy decision. Budget below it is usually a waste, and the useful thing to know is which side of the line you are on.

The budget floor formula

Worked once, slowly: if your landing page converts 5% of clicks and your keywords cost $4 a click, you need 600 clicks to produce 30 conversions, and 600 clicks at $4 is $2,400 a month. That is your floor. At $800 a month you would generate roughly 10 conversions, the bidding algorithm never leaves the learning phase, and your cost per acquisition stays high and volatile no matter how well the account is managed.

Why 30 conversions, and not some other number

This is the part most budget guides skip, and it is the only part that makes the formula defensible rather than arbitrary.

Google's own documentation sets a hard minimum of 15 conversions in the trailing 30 days to use Target CPA bidding, and recommends at least 30 in that window for the strategy to perform well. The gap between those two numbers is where a lot of small accounts live and quietly underperform: technically eligible, practically starved. Smart Bidding sets a bid for every individual auction based on signals it has learned from your conversion history. With ten data points a month it is guessing; with thirty it is estimating; practitioners generally find genuine stability higher still, around 50 or more a month, particularly in lead generation where conversion quality varies between one form fill and the next.

So 30 is deliberately the recommended threshold rather than the eligibility minimum or the comfortable target. It is the lowest number at which the machinery you are paying for actually starts working. Use 50 in the formula instead if you want the stable-performance floor rather than the functional one.

Floors worked out with 2026 benchmarks

The table below applies the formula using published 2026 cost-per-click benchmarks. The conversion rates are stated bands, not precise industry figures, because conversion rate varies far more by landing page quality than by sector — substitute your own the moment you have 30 days of data.

Two things jump out of that table. E-commerce has the lowest floor despite the worst conversion rate, because clicks are cheap — which is why "what is a good conversion rate" is the wrong benchmark to manage against on its own. And the two home services rows differ only in landing page quality, and the floor doubles. Hold that thought.

Why published CPC averages disagree by 80%

You will find 2026 sources quoting an all-industry average search CPC of $5.42 and others quoting $2.96 for the same year. Neither is lying. The spread comes from three things, and knowing which you are reading matters more than which number you pick.

• Account mix. Datasets built from agency-managed accounts skew toward higher-budget, more competitive verticals than datasets drawn from platform-wide or self-serve accounts.
• Network scope. Search-only averages run far above figures that blend in Display, Demand Gen or Shopping inventory.
• Measurement window. Benchmark studies typically cover a trailing twelve months, so a study published in 2026 may be describing mostly 2025 auctions. Cross-industry search CPC rose roughly 12% year on year into 2026, the steepest annual rise since 2021, so a stale window understates today's costs.

Practical rule: use a benchmark only to sanity-check a floor before you have data, and replace it with your own trailing 30-day CPC as soon as you have one. If you must pick blind, pick the higher figure and treat the lower one as your optimistic case.

The cheapest way to lower your floor is not your bidding

Look at the formula again. Cost per click is set by an auction you only partly control. The conversion-rate term is sitting in the denominator, entirely inside your own website, and it moves the floor proportionally.

Moving a landing page from 2% to 5% does not make your ads 2.5 times better. It makes a budget that could not previously work, work. That is a different and much more valuable claim, and it is why the sequence matters: fix the page, then set the budget, then optimise the account. Doing it in the other order is how businesses conclude that "Google Ads does not work for us" after spending $6,000 finding out.

This is not a hypothetical lever in 2026 specifically. Benchmark data for the year shows click-through rates rising roughly 7.5% while conversion rates fell in 13 of 14 industries measured. Ads are getting better at getting the click; pages are getting worse at converting it. The constraint has moved off the ad and onto the page it lands on.

What to do when the floor is above your budget

Most small businesses reading this will calculate a floor higher than the number they had in mind. That is useful information, not a dead end. There are five honest responses, in rough order of preference.

1. Shrink the market until the floor fits. The floor scales with the volume you are chasing, not the size of your ambition. One city instead of a region, one service line instead of five, business hours only. A budget that is inadequate across a state can be comfortably above the floor in a single postcode.
2. Cut to only the highest-intent keywords. Exact-match terms with obvious purchase intent convert several times better than broad research terms, which raises the CVR term and lowers the floor. Fewer, better keywords beats broad coverage on a thin budget every time.
3. Raise the conversion rate before raising the budget. Per the table above, this is usually the cheapest available move, and it improves every other channel at the same time.
4. Change the bid strategy to match the data you have. If you genuinely cannot reach 30 conversions a month, Smart Bidding strategies that depend on conversion volume are the wrong tool. Manual or click-focused bidding with tight keyword control is less glamorous and more appropriate below the floor.
5. Do not run Search yet. If the floor for a viable Search campaign is $4,000 and you have $900, that $900 spent on organic, local listings or a better site will return more than a starved campaign will. This is a legitimate answer and almost no agency will give it to you.

The soft conversion trap

There is a sixth response, it is extremely common, and it is a mistake: reach 30 conversions by counting easier things. Newsletter signups, PDF downloads, page views on a contact page, clicks on a phone number that nobody answered.

It works arithmetically and fails commercially. Smart Bidding optimises toward whatever you tell it is a conversion, so an account trained on soft conversions gets very good at producing soft conversions. Your conversion count goes up, your reported cost per conversion goes down, and your actual booked revenue does not move. This is the mechanism behind the pattern where cost per qualified lead rises while cost per lead falls — and it is much harder to unwind after six months of training than it is to avoid at setup.

If you must use a softer conversion action to reach volume, choose one that correlates with revenue, such as a booking page reaching the confirmation step, and keep a separate view of your genuine sales so you can see the two diverge.

Ad spend is not your budget

One last correction that changes the number materially. Everything above calculates media spend. Your actual monthly cost includes management, whether that is an agency retainer, a freelancer, or your own time. Agency fees commonly run as a percentage of spend or a flat retainer, and on small accounts the retainer is frequently larger than the ad spend itself.

That produces a specific trap: an account at $1,200 a month in media with a $1,500 retainer is spending $2,700 to buy $1,200 of learning-phase data. If your total budget is around that level, the better allocation is usually more media and less management — a simpler account, fewer campaigns, a well-built landing page and light oversight. Sophisticated account management earns its fee above the floor, where there is enough data for the decisions to matter. Campaign type selection, audience layering and bid experimentation are all downstream of having enough conversions to read.

Once you are comfortably above the floor, the question changes from "is this enough" to "what does the next dollar cost", which is a genuinely different discipline — and the point at which scaling without wrecking your CAC becomes the thing to worry about.

Frequently asked questions

How much should a small business spend on Google Ads per month in 2026?

Rather than a fixed range, calculate your floor: 30 divided by your conversion rate, multiplied by your cost per click. Thirty conversions a month is the volume Google recommends for Smart Bidding to perform well. Using 2026 benchmarks, that puts the floor near 1,530 dollars a month for home services at a 10 percent conversion rate, around 3,375 dollars for legal services, and around 1,242 dollars for e-commerce. Budget above your floor is a strategy choice; budget below it usually buys data the system cannot learn from.

What is the minimum budget for Google Ads to work?

There is no universal dollar minimum, but there is a mechanical one. Google requires at least 15 conversions in the trailing 30 days to use Target CPA bidding and recommends at least 30 for it to perform well. Your minimum budget is whatever it costs to buy that many conversions in your market, which depends on your cost per click and your landing page conversion rate. If you cannot reach it, narrow your geography or keywords until you can, or use bidding that does not depend on conversion volume.

Why do published Google Ads CPC benchmarks disagree so much?

Because they measure different things. Datasets built from agency-managed accounts skew toward more competitive verticals than platform-wide data, search-only averages run well above figures that blend in Display or Shopping inventory, and most benchmark studies cover a trailing twelve months, so a 2026 report may largely describe 2025 auctions. With cross-industry search CPC up roughly 12 percent year on year, a stale window understates current costs. Use benchmarks only until you have your own trailing 30-day CPC.

Does increasing my landing page conversion rate really lower my required budget?

Yes, and proportionally. Conversion rate sits in the denominator of the floor formula, so doubling it halves the clicks you need and therefore halves the floor. At a 5 dollar cost per click, a page converting at 2 percent needs about 7,500 dollars a month to reach 30 conversions, while the same page at 5 percent needs about 3,000. This matters more in 2026 than usual: benchmark data shows click-through rates rose about 7.5 percent while conversion rates fell in 13 of 14 industries measured.

Should I count newsletter signups as conversions to hit the 30 conversion threshold?

Generally no. Smart Bidding optimises toward whatever you define as a conversion, so an account trained on soft actions becomes efficient at producing soft actions while booked revenue stays flat. Your cost per conversion falls and your cost per qualified lead rises. If you need a softer action to reach volume, pick one that correlates with revenue, such as reaching a booking confirmation step, and track genuine sales separately so you can see if the two diverge.

Is my agency retainer part of my Google Ads budget?

It is part of your total cost, and on small accounts it is often the larger half. An account running 1,200 dollars of media under a 1,500 dollar retainer spends 2,700 dollars to buy learning-phase data. Below the floor, the better allocation is usually more media and less management, with a simple account structure and a strong landing page. Detailed account management earns its fee above the floor, where there is enough conversion data for optimisation decisions to be readable.

How long before a Google Ads campaign at the right budget shows results?

Smart Bidding needs a trailing 30-day conversion window to work from, so plan on at least two full months before performance data means anything, and expect the first few weeks to look worse than the eventual steady state while the learning phase runs. Changing targets, budgets or conversion actions restarts that process, which is why under-budgeted accounts that get frequently adjusted never stabilise.

The takeaway

The budget question has a real answer, and it is specific to you rather than to your industry. Take your cost per click, take your landing page conversion rate, and work out what it costs to buy 30 conversions a month. If that number is below your budget, you have a strategy question about how much further to go. If it is above, you have a much more useful problem than you had five minutes ago: you now know that narrowing your geography, tightening your keywords or fixing your landing page is worth more than any amount of careful bid management. The businesses that waste money on Google Ads are rarely the ones that spent too much. They are the ones that spent just enough to generate data and never enough for anything to learn from it.