Google Ads budget, how a daily figure is actually spent, what the forecast tool is modelling and the smallest test that teaches you anything

Google Ads has no minimum spend, no account fee and no rate card. You set an average daily budget, you are billed for what the auction charges, and the account can be run on almost any number. That freedom is the problem, because a budget too small to produce a readable result costs the same attention as one that works and teaches you nothing. The question worth asking is not what Google requires. It is what budget is large enough to produce enough conversions to decide something.

Verified against Google's India geo target alone on 22 October 2026, google ads cost estimator draws 90 searches a month in India, with a further 50 a month in India on the same reading for google ads minimum budget. This page is about sizing and pacing a budget. For what sets the price of each click and the costs beyond the clicks, our page on Google Ads PPC cost, what you pay and what sets the price does that and this page builds on it rather than repeating it.

How a daily budget is actually spent

The figure you enter is an average, not a cap, and understanding that single fact removes most of the surprise from a first invoice.

• Any single day can exceed your daily figure. Google may spend more on a day when there is more opportunity, and less on a quiet day, to even out over the month.
• The month is what is capped. Google documents the limit as the average daily budget multiplied by 30.4, the average number of days in a month. Read the current wording on Google's own help documentation when a number matters, because this is the kind of detail that gets revised.
• A campaign paused mid month does not get its unspent allowance back. The averaging works forward from the budget in place, so turning campaigns on and off repeatedly makes the monthly total hard to predict for a reason that has nothing to do with the auction.
• Budgets are set per campaign, or shared across several. A shared budget will be consumed by whichever campaign has the most opportunity, which is efficient and also means a campaign you care about can be starved by one you care less about.

The practical consequence is that a daily figure should be read as roughly one thirtieth of what you are agreeing to spend in a month, and the monthly number is the one to put in a plan.

Sizing the budget from your own numbers

Work it out in four steps rather than picking a round figure. Replace every number here with your own.

• Start from what a customer is worth. Take your gross margin on one customer. If you earn 5,000 of margin on an average customer, that is your ceiling for winning one.
• Divide by your conversion rates to get a workable click price. If one visitor in fifty becomes an enquiry and one enquiry in four becomes a customer, you need two hundred clicks a customer. A 5,000 ceiling across two hundred clicks is 25 a click, and that is the price above which the account stops working.
• Decide how many conversions you need to learn anything. This is the step people skip. A handful of conversions is not a result you can act on; it is noise with a direction. Set the test at a number of conversions you would be willing to make a decision on, and work backwards.
• Multiply. Conversions wanted, times clicks per conversion, times the click price you can afford, gives the test budget. If that number is larger than you can spend, the honest options are to narrow the targeting until the click price falls, or to improve the conversion rate first, which costs no media at all.

That arithmetic also answers the minimum budget question properly. The minimum is not a Google rule. It is whatever produces enough conversions to decide, and it is specific to your margin and your rates.

What the forecast tool is modelling, and what it leaves out

Google's planning tools will forecast clicks, impressions and cost for a keyword set at a bid you choose. The forecast is a model and it is useful, and three things it does not include account for most of the gap between a forecast and an invoice.

• It assumes you get the impressions. A forecast is built on the auction as it sees it, not on your account's history, your relevance or your landing pages, so a new account commonly wins fewer impressions than the model implies.
• It forecasts the keywords you gave it. Actual spend also goes to the variations your match types allow in, and that set is wider than the list you typed unless your negatives are tight.
• It has no view of your conversion rate. A forecast of clicks is not a forecast of customers, and the step between the two is the one that decides whether the plan works.

Use the forecast for the shape of the thing: whether a keyword set has enough volume to be worth building, and roughly what order of click price to expect. Use your own account data for anything you will commit money against.

Where a small budget leaks

Four leaks account for most of a wasted small budget, and all four are settings rather than strategy.

• Broad match without negatives. The widest match types will find you queries you never intended, and on a small budget a handful of irrelevant clicks a day is a large share of the whole. Start tighter than feels comfortable and widen with evidence.
• The search partner network and display left on inside a search campaign. These are different inventory with different behaviour. Separate them so you can read them separately, rather than averaging two things into one number.
• Locations and times nobody buys in. If your business serves three cities and answers the phone during office hours, a budget running nationally and overnight is spending money on impossibilities.
• No conversion tracking. Without it, every optimisation is a guess and automated bidding has nothing true to aim at. This is the single highest return hour of work in a new account.

Pacing, and the mistake of judging week one

A new campaign does not spend evenly or perform representatively in its first days. Prices and placements settle, automated strategies collect data, and early readings are both volatile and systematically unlike the steady state.

So set the window before you start. Decide in advance how long the test runs and how many conversions it has to produce, write both down, and resist changing targeting or bids inside the window unless something is clearly broken. A campaign altered every second day has no window at all, and the account becomes a sequence of opinions rather than a measurement.

When you do compare, compare like with like: the same locations, the same match types, the same conversion definition. Changing the conversion action mid test is the quiet version of this mistake and it invalidates everything before it.

Questions about setting a Google Ads budget

Is there a minimum spend for Google Ads

No. There is no required minimum and no account fee. The practical minimum is whatever produces enough conversions to make a decision, which comes out of your own margin and your own conversion rates rather than out of a Google rule.

Why did Google spend more than my daily budget yesterday

Because the daily figure is an average rather than a cap. Google can spend more on a day with more opportunity and less on a quieter day, with the total limited over the month to the average daily budget multiplied by 30.4, which Google documents on its own help pages.

How do I estimate a monthly cost before starting

Use the forecast tool for the order of magnitude and your own arithmetic for the commitment. Margin on a customer, divided by clicks needed per customer, gives an affordable click price; multiply by the clicks your test needs and you have the monthly figure.

Should I set one budget or several

Separate budgets per campaign when the campaigns matter differently to you, because a shared budget flows to whichever campaign finds the most opportunity. Shared budgets are convenient and they remove your ability to protect a priority.

How long before I can judge the result

Long enough to collect the number of conversions you decided in advance would be enough to act on. Set that figure and the window before launch, because judging early and changing settings weekly produces an account with no readable history.

What is the first thing to fix on a small budget

Conversion tracking, then negative keywords, then locations and schedule. Those three remove most of the waste and they cost nothing but an afternoon.

The enquiries with no campaign behind them

Every figure on this page belongs to a system that counts what it sold you: impressions, clicks, conversions, spend. A share of your enquiries now arrives from a process that counts nothing. Somebody asks an assistant which supplier to use, the answer names two or three businesses, and the person contacts one of them directly. There is no campaign, no keyword and no cost per click, so that demand is invisible in the account and easy to credit to whatever was running at the time. Knowing whether you are in those answers is a measurement rather than a guess, and it changes how you read a paid report. AI Knows Us takes that reading across ChatGPT, Gemini, Perplexity, Claude and Google AI Overviews, with the sources behind each answer, so the enquiries your ad account cannot explain stop being a mystery.