The Google Ads budget floor: what you have to spend before the algorithm can do its job

Ask how much you should spend on Google Ads and you get a range. Ranges are useless for deciding whether your number is viable. There is an arithmetic answer instead, built from two published benchmarks and one requirement Google states openly, and it explains why identical budgets succeed in one category and fail in another.

Why "how much should I spend" is the wrong first question

Every answer you will find to that question is a range: $1,000 to $2,500 a month, more if you are in a competitive category. The range is not wrong. It is just unexplained, which makes it useless for deciding whether your $900 is a reasonable start or money set on fire.

There is a better question with an arithmetic answer. Google's Smart Bidding strategies need a minimum flow of conversion data to work: Google's own guidance is at least 30 conversions in the past 30 days at campaign level before Target CPA behaves predictably, and Target ROAS is usually quoted higher. Below that, the bidding algorithm is guessing, every significant change restarts a one-to-two week learning period, and you cannot tell a bad campaign from an unlucky fortnight.

So the floor is not a matter of taste. It is whatever it costs, in your market, to buy 30 conversions a month. Everything else follows from that.

The budget floor formula

A quick sanity check on the formula, using published 2026 data. WordStream by LocaliQ's 2026 Search Advertising Benchmarks report, based on more than 13,000 US campaigns running between April 2025 and March 2026, puts the all-industry average CPC at $5.42 and the all-industry average conversion rate at 8.18%. Divide one by the other and you get a cost per lead of $66.26. The same report's published average cost per lead is $66.69. The formula reproduces the observed number to within one percent, which is what you want from a model before you trust it with your budget.

Multiply that $66 by 30 and the all-industry Google Ads budget floor comes out at roughly $2,000 a month. That is where the familiar "$1,000 to $2,500" advice comes from, even though almost nobody states the derivation. It also explains why the advice fails so often: the average is an average, and the spread underneath it is enormous.

The floor by industry

Below, the published 2026 average CPC for each category, converted into a monthly floor. Because per-industry conversion rates are not published for every category in the same dataset, the table shows two columns: the floor if your landing page converts at the all-industry average of 8.18%, and the floor if it converts at 4%, which is a common reality for a service business sending paid traffic to a general page.

Two readings of that table are worth making explicit. First, the floor varies by a factor of more than twelve across ordinary business categories. A restaurant group and a personal injury firm are not in the same conversation about minimum budgets, and any advice that gives them the same range is not advice. Second, the conversion rate column moves the floor more violently than the CPC does. Halving your conversion rate roughly doubles the money you need before Google Ads can optimise at all.

Conversion rate is the cheapest lever on your floor

Hold CPC at the all-industry $5.42 and vary only the conversion rate, using published per-industry conversion rates from the same 2026 dataset:

Same clicks, same prices, and the entry ticket moves from $1,000 to over $6,000 purely on what happens after the click. This is the part of the budget conversation that gets skipped, because the budget conversation happens with a media buyer and the conversion rate lives on a page somebody else built.

The practical consequence: if your floor is above your budget, fixing the landing page is usually cheaper and faster than finding more money. A move from 3% to 6% halves the floor. That is a page-level project, not a media-budget project, and we have written about what a good landing page conversion rate looks like by industry if you want to know whether yours has room.

The floor is per campaign, not per account

This is where most under-funded accounts actually break. The 30-conversion guidance is campaign-level. A $2,000 monthly budget in one campaign can clear the all-industry floor. The same $2,000 split across five campaigns — brand, non-brand, a competitor campaign, Performance Max and a remarketing campaign — clears the floor in none of them, and now you have five campaigns all permanently in learning.

Consolidation is the fix nobody likes, because the account looks less sophisticated afterwards. Fewer campaigns, fewer ad groups, broader match types with tight negatives, one clear conversion action. The structural instinct that made sense under manual bidding actively harms automated bidding, and this is the same underlying tension as the Performance Max versus Search campaign question.

If your budget is below the floor

Four options, in the order we would try them. Only the last one involves giving up.

1. Shrink the market, not the budget. The floor is a function of what a click costs in the market you are bidding into. Cut the radius to the area you actually serve, run only during hours you can answer the phone, and restrict to the bottom-funnel terms where intent is explicit. A smaller market with the same money raises your effective spend per opportunity. This is the single most effective move for a local service business.
2. Raise the conversion rate. As above: it moves the floor faster than anything else you control. A dedicated landing page per service instead of a shared homepage, a form that asks three questions instead of nine, a visible phone number, an answer to the price question.
3. Count a conversion that happens more often. If a booked consultation happens eight times a month, Smart Bidding cannot learn from it. A laddered setup — optimise toward a more frequent upstream action such as a qualified call start or a quote request, while still reporting on the downstream sale — can get the campaign above 30 events. Do this carefully: optimise toward something too easy and you will buy volumes of junk, which is the failure mode we set out in the cost per qualified lead model.
4. Accept that this is not your channel this quarter. If a legal firm's floor is $7,400 and the budget is $1,500, no amount of structure fixes that gap. The honest recommendation is local SEO, a Google Business Profile, referral work or a narrower paid channel until the budget or the conversion rate changes. An agency that takes the $1,500 and promises to make Google Ads work anyway is selling you twelve months of learning period.

Learning budget versus steady-state budget

One more distinction that saves arguments in month two. The floor is the learning budget: the spend required for the system to reach a stable state. It is not a performance target, and the early weeks will look worse than the average, because they are the average's raw material.

Budget for the floor, hold it steady for at least six to eight weeks, and resist mid-flight budget changes. Significant changes — a bid strategy switch, a target adjustment, a budget change of roughly 20% or more — push the campaign back into a learning period of one to two weeks. A nervous client who cuts the budget in week three, restores it in week five and raises the target in week six has bought three learning periods and zero stable data, and will conclude that Google Ads does not work for their business.

Once the campaign is stable and profitable, the constraint changes completely: the question stops being "what is the minimum" and becomes "how far can this scale before CAC degrades", which is a different problem with a different playbook.

What the floor is not

It is not a prediction of profitability. Clearing the floor buys you a campaign that can learn; whether that campaign makes money depends on your margins and lifetime value, which is what ROAS and CAC benchmarks are for. It is entirely possible to clear the floor and still find that the maths does not work, and it is much better to discover that with clean data than with a campaign that never had enough signal to tell you anything. That is the first thing we establish on any paid media engagement, before a single campaign is built.

It is also not fixed. Published CPCs moved sharply over the last year — real estate up 27.3%, personal services and health and fitness up 23.4% — in categories where AI Overviews have reduced organic click-through and pushed more businesses into paid results. A floor calculated eighteen months ago is not today's floor, and in the categories most exposed to AI search, it has risen faster than most budgets have.

Frequently asked questions

What is the minimum budget for Google Ads in 2026?

There is no platform minimum, but there is a practical one: enough spend to produce about 30 conversions per campaign per 30 days, which is Google's stated guidance for Smart Bidding to optimise reliably. Using 2026 published benchmarks of a $5.42 average CPC and an 8.18% average conversion rate, that works out at roughly $2,000 a month for an average business, and considerably more in expensive categories such as legal or home improvement.

How do I calculate my own Google Ads budget floor?

Multiply 30 by your average CPC divided by your landing page conversion rate. For example, a $6 CPC and a 5% conversion rate gives a $120 cost per conversion and a $3,600 monthly floor. Use your own account's numbers if you have them; use published category benchmarks if you are starting from scratch, then recalculate after 60 days of real data.

Is $500 a month enough for Google Ads?

Only in a low-cost category with a strong landing page and a tightly restricted market. At an all-industry $5.42 CPC, $500 buys roughly 92 clicks a month, which at an 8.18% conversion rate is about 7 conversions, well below the level Smart Bidding needs. It can work for a local business bidding on a handful of bottom-funnel terms in a small radius with a page converting above 10%, and it will not work for a broad campaign in a competitive category.

Why does Google Ads need 30 conversions a month?

Smart Bidding strategies set bids by predicting conversion probability for each auction, and those predictions need recent examples. Google's guidance is at least 30 conversions in the past 30 days at campaign level for Target CPA. Below that threshold the model has too little signal, performance is volatile, and you cannot distinguish a genuine trend from noise.

Does the 30-conversion threshold apply per campaign or per account?

Per campaign. This is why splitting a modest budget across five campaigns is worse than putting it into one. A $2,000 monthly budget can clear the floor in a single consolidated campaign and clears it in none of five, leaving every campaign permanently in a learning state.

What should I do if my budget is below the floor?

In order: shrink the market rather than the budget (tighter geography, hours and bottom-funnel keywords only), raise the landing page conversion rate, which moves the floor faster than anything else, or optimise toward a more frequent upstream conversion while still reporting on the sale. If the gap is still large after all three, Google Ads Search is not the right channel this quarter and the budget is better spent on local SEO or a narrower channel.

How long should I leave a budget unchanged?

At least six to eight weeks once the campaign is live. Significant changes, including budget changes of roughly 20% or more, bid strategy switches and target adjustments, restart a learning period that typically lasts one to two weeks. Repeated mid-flight changes produce a sequence of learning periods and no stable data to judge.

The takeaway

Do the arithmetic before the negotiation. Thirty conversions times your cost per conversion is the number that decides whether a Google Ads plan is a plan or a hope, and it is the number that explains why the same $1,500 is generous for a restaurant group and hopeless for a law firm. If your floor is above your budget, your cheapest move is almost always the landing page rather than the media budget, because conversion rate moves the floor twice as hard as CPC does. And if the gap survives every lever, say so: the most valuable thing an honest media plan does is tell you which quarter not to start.