Google Ads Budget: How to Size, Split and Change It

You size a store’s Google Ads budget from three numbers: the most you can pay for one order, the conversions each campaign needs to learn, and how much demand is still left to buy. Split the money into separate campaigns only when product groups earn differently and each part keeps enough conversions. The GetProfit portal works with budget steps of up to 20% and judges each one after two weeks. In GetProfit data on 1,360 store-months, months with higher spend brought more revenue at roughly unchanged ROAS.

Budget is one of two levers behind ad revenue. The revenue your ads bring in equals your ad spend times the return on each unit of it. So you grow ad revenue either by putting in more money or by getting a better return on it. This guide covers the money side: how much to spend, where it goes and how to move it. For how the two levers work together, see growing a store with Google Ads.

What does a Google Ads budget actually limit?

In Google Ads you set an average daily budget for each campaign. It works as an average over the month. According to Google Ads Help (About average daily budgets), Google spends more on days when clicks and conversions are more likely and less on quieter days. Two limits hold the total:

• Daily: for most campaigns, no more than twice the average daily budget on any one day.
• Monthly: for most campaigns, no more than 30.4 times the average daily budget in a calendar month. 30.4 is the average number of days in a month.

To turn a monthly plan into a daily budget, divide it by 30.4. On rare days the clicks can cost more than the daily limit. Google’s Budgets overview explains that you pay only up to the limit and Google covers the difference. For why a single day can run up to twice the budget, see the daily budget explained.

Google offers three types of budget, and most stores need only the first.

Google sets out the rules for one fixed sum over set dates on its page about campaign total budgets. One budget spread across several campaigns has its own page, shared budgets. A campaign total budget keeps re-pacing itself. If the campaign spends slowly in the first days, Google raises daily spend for the days left, so the campaign uses the whole sum by the end date (How budget changes take effect).

How much should an online store spend on Google Ads?

The right amount follows from your margin and your conversion volume. Work it out in four steps.

1. Find the most one order can cost. Average order value × gross margin gives the ad cost at which an order breaks even, before your other costs. The same limit in ROAS terms is 1 ÷ gross margin. Keep this number at hand. Your ad account has no purchase prices, so the ROAS it reports is based on revenue, not profit.
2. Count the conversions each campaign needs. A bid strategy goes through a learning period before it calibrates to a new goal. According to Google’s help page Duration of the learning period for campaigns and what affects it, that can take up to around 50 conversions, or 3 conversion cycles. The portal treats 30 conversions a month per campaign as the minimum and 50 or more as comfortable.
3. Give each campaign enough money per day. For Performance Max, Google’s Ads API documentation (Create a Performance Max Campaign Budget) advises an average daily budget of at least three times your cost per conversion. A lower budget can mean a slower ramp-up and fewer conversions. The portal uses the same floor: three times the target cost per conversion, per day.
4. Check how much demand is left. The Search lost IS (budget) column shows how often your ads missed a search because the budget ran out. If it is close to zero, something other than the budget holds the campaign back. Its pair, Search lost IS (rank), shows the share lost to low Ad Rank.

Example store, not client data.

A tableware store with 3,000 products spends 40,000 a month, takes 300 orders worth 180,000 and keeps a 35% gross margin. Its average order value is 600, so an order breaks even on ads at 600 × 35% = 210. Today its cost per conversion is 40,000 ÷ 300 ≈ 133. Its ROAS of 4.5 sits well above its break-even ROAS of 1 ÷ 0.35 ≈ 2.86.

Three times the cost per conversion is about 400 a day for each Performance Max campaign. The monthly budget of 40,000 is about 1,316 a day (40,000 ÷ 30.4). That covers three campaigns at the floor, 1,200 a day; a fourth would need 1,600. The 300 orders a month could support six campaigns at 50 orders each, so here the budget sets the limit: three campaigns.

The budget calculator runs the same arithmetic on your own average order value, margin and conversion rate. For what happens below the floor, and how small a campaign can be while it still learns, see the minimum budget for Performance Max.

How should you split the budget across campaigns?

For most stores the split happens inside Performance Max. In GetProfit data, Performance Max runs in 141 of 146 stores and takes a median 95.7% of their ad budget (June 2025 – June 2026). In the 81 stores that also run Search, Search takes a median 5.3% of the budget; in the 74 that run Shopping campaigns, Shopping takes a median 6.5%. So the practical question is how to divide money between several Performance Max campaigns, with a small Search campaign next to them.

A split pays off when the product groups behave differently and each part still gets enough conversions to learn. The portal and Google set these rules:

Once the campaigns exist, compare each one’s share of spend with its share of revenue. A campaign whose share of spend is bigger than its share of revenue is the first place to trim. One whose share of revenue is bigger is the first place to add money. The full method is in allocating budget across categories and brands.

More budget, same median ROAS: what 1,360 store-months show

We compared 1,360 store-months (one store’s results in one calendar month) from 144 stores between June 2025 and June 2026. For each month we took the change in a store’s spend, revenue, ROAS and cost per click against the month before. From each change we subtracted the median change of all stores in the same calendar month. That removes the season and platform-wide shifts and leaves what a store did differently from its peers.

Medians, relative to the median store in the same month. “ROAS held” is the share of store-months where ROAS fell by no more than 5% against peers. GetProfit data, 144 stores, June 2025 – June 2026.

What the table shows:

• Revenue followed spend, and ROAS barely moved. In every band the median ROAS changed by less than 4%. The rank correlation between the change in spend and the change in ROAS is +0.026, practically zero.
• Clicks got more expensive as spend grew. In months where spend rose more than 70% above peers, the cost per click went up by a median 9.2%. ROAS still held in the median month: revenue grew along with the volume.
• Cuts came with lower revenue, and ROAS stayed flat or slipped. In months with a cut of 10–30%, ROAS stayed where it was and revenue fell by a median 18.9%.
• Months with raised spend went two ways. Of 381 months where spend rose more than 20% above peers, 227 (60%) held ROAS, with a median revenue gain of 93.1%. In the other 154 (40%) ROAS fell, and revenue grew by a median 19.0%.

This is an observation, not an experiment. A store whose sales were already rising may have added money because of that, so the link can run in the other direction. We count ROAS here on revenue: we have no purchase prices, so these numbers say nothing about profit.

Monthly ROAS swings a lot anyway. A typical store’s monthly ROAS sits 16% away from its own median (GetProfit data, 110 stores, June 2025 – June 2026). One weak month after a budget step proves little on its own. For how to raise the budget in stages, and what to check after each stage, see scaling Google Ads without breaking ROAS.

How do you change the budget without breaking ROAS?

The portal works by a short set of rules for budget edits:

1. Move in steps of up to 20%. The portal treats ±20% at a time as the working step for a budget. For a target ROAS the step is ±15%, no more often than once every one to two weeks. In the portal’s change log, a budget step above 50% counts as sharp.
2. Change one thing at a time. If you change the budget, the target and the campaign structure in the same week, nobody can tell which of them moved the result.
3. Wait two weeks before the verdict. The portal gives a first conclusion on a change after a week and a confident one after two.
4. Expect pricier clicks. In our data the cost per click rose by a median 3.4% when spend went up 10–30% above peers, and by 9.2% when it went up more than 70%.
5. Know what a mid-month change does to the limits. After a change, Google caps the rest of the month at what you have already spent plus the new daily budget times the days left. On the day of the change, the daily limit follows the highest budget you set that day (How budget changes take effect).

Google’s help lists what puts a bid strategy into “Learning” status: a new or reactivated strategy, a changed strategy setting, and campaigns, ad groups or keywords added or removed (About bid strategy statuses). Budget edits are not on the list, though they still change how much traffic the campaign buys. After a change from the list, results can swing during the learning period. Google says you may not want to measure performance until it is over.

Campaigns on a target ROAS need one extra check. Since 17 August 2026, a budget-limited campaign with a target-based strategy performs more consistently towards its target. If it used to beat the target, it now delivers closer to it (Changes to target based bid strategies).

Google leaves it to you to change targets and budgets, and it points to a higher budget as the way to get more volume at the target you set. What this means for a store, with worked numbers, is in Google’s August change to target-based bidding.

The portal’s change log records budget edits, works out what things looked like before and after, and says whether each change worked or hurt.

Is your ceiling the budget or Ad Rank?

Impression share shows whether more money would buy more demand. Google describes it as a way to see whether your ads might reach more people if you raise your bid or your budget (About impression share).

Google’s help on these columns (Get impression share data) defines two losses of impression share, or IS. Search lost IS (budget) is the share of time your ads did not show because the budget was too small. Search lost IS (rank) is the share lost to low Ad Rank. The first loss points at the budget; the second points at bids and ad quality.

In GetProfit data, rank is the ceiling far more often than budget. Across 125 stores with at least 120 days of data (June 2025 – June 2026), the median store won 45.8% of search impressions. The median shares lost were 50.2% to rank and 13.6% to budget. In 110 of the 119 stores where we could measure both losses, rank cost more impressions than budget.

Impression share has limits. It describes a store’s own eligible pool, which its catalogue and budget have already narrowed, and does not show the size of the market. For Performance Max the metric is incomplete: Google’s About impression share page says it counts only Search and Shopping impressions. In our data it is missing for Demand Gen, Display and Video campaigns altogether.

Google marks a campaign “Limited by budget” in budget pacing insights when it missed 5% or more of its potential traffic in the past week (About budget pacing insights). Google bases its recommended budget on recent performance, the current budget, keywords and targeting. If you rarely reach your daily budget, it shows no recommendation (Fix “Limited by budget” status).

Whether to add money or loosen the target in that case is covered in the guide to ‘Limited by budget’. For both losses side by side, with what to fix first, see lost impression share: budget vs rank.

How do you pace a budget through the month?

Pacing means checking mid-month whether each campaign is spending its monthly limit too fast, too slowly or on track. Google gives two tools for it, and they cover different campaigns.

According to Google’s help page About your budget report, Performance Max campaigns are currently not compatible with the budget report. Stores that run mostly on Performance Max need the insights page or their own sheet. A monthly pacing sheet for Performance Max does the check by hand: spend to date, divided by days passed, times the days in the month, compared with the monthly limit.

An ad schedule leaves the monthly limit as it is. With ads off on some days, the limit stays at 30.4 times the daily budget, and Google paces towards it on the days the ads run. On any active day, spend still stays within twice the daily budget (About spending limits). When weekends eat the budget or a prepaid balance runs out before the top-up, see weekend overspend and empty prepaid balances.

How should the budget follow the season?

In GetProfit data on 96 stores (July 2025 – June 2026), the three best months brought a median 2.62 times the ad revenue of the three worst. In 92 of the 96 stores the gap was at least 1.5 times, so almost every store has a season. For the median store, the shortfall against its own median month cost 10.8% of its yearly ad revenue. To plan the budget around seasonality, start from these two numbers for your own store.

The budget usually falls with demand. In 61 of the 96 stores (64%), spend in the three worst months was below 70% of spend in the three best. ROAS does drop in the trough: the median store’s ROAS there was 0.70 of its peak ROAS, so part of the fall is real demand. Our data can’t separate how much the budget cut adds on top of that.

For the run-up to a peak, the portal works by these rules:

• raise the budget by 20–30% about two weeks before the season;
• lower the target ROAS by 10–15% at the same time;
• hold off raising the target ROAS until the peak is over.

The pre-season step is the one planned move that can exceed the usual 20% budget step: you make it before demand rises, not in reaction to a result.

For short sales, Google offers seasonality adjustments, which tell Smart Bidding (the strategies where Google picks the bid in every auction) to expect a change in conversion rate. Google recommends them for events of 1 to 7 days and only for major changes, since Smart Bidding already manages seasonal events (About seasonality adjustments). For turning your own seasonal curve into a budget for each month, see a month-by-month Google Ads budget plan.

What to do this month

1. Write down your break-even cost per conversion: average order value × gross margin. Keep every target you set on the profitable side of it.
2. Count conversions per campaign over the last 30 days. Campaigns with fewer than 30 a month are candidates for merging rather than for separate budgets.
3. Check every Performance Max daily budget against three times its cost per conversion. Below that floor, Google warns that the ramp-up may be slower and conversions fewer.
4. Add the Search lost IS (budget) and Search lost IS (rank) columns. They show whether the budget or Ad Rank is your ceiling before you add money or raise bids.
5. Compare each campaign’s share of spend with its share of revenue. The gap shows where the next budget step should go.
6. Make one change of up to 20%, note the date and leave it for two weeks. A clean before-and-after shows what the change did.
7. Review targets on budget-limited campaigns with a target ROAS. Since 17 August 2026 they deliver closer to the target, so a target below the campaign’s current results pulls them down to it.
8. Put the pre-season budget step in the calendar two weeks before your peak. The portal’s rules place it before the season and hold off raising the target ROAS until the peak is over.

Who changed what in your ads — and how it ended. The portal keeps a log of changes: budgets, bids, conversion goals, product sets, whole campaigns. For each one it works out what things were like before and after. The portal changes nothing without your consent.

See what changed in your account →

Frequently asked questions

Can Google Ads spend more than my monthly budget?

No: for most campaigns you pay at most 30.4 times the average daily budget in a calendar month. Google can occasionally serve ads worth more than that, and then it covers the difference, as its page on spending limits explains. A campaign total budget is a hard cap for its whole period.

Can Performance Max share a budget with other campaigns?

No. Google’s help lists shared budgets for Search, Shopping, Display and Video campaigns only. The Google Ads API documentation also states that a Performance Max budget can’t be shared, so each Performance Max campaign needs its own budget.

Does a budget change restart learning?

Not on Google’s list: the triggers its help names for the “Learning” status are a new or reactivated strategy, a changed setting, and added or removed campaigns, ad groups or keywords. A big budget step still changes the traffic the campaign buys. That is why the portal works with budget steps of up to 20% and judges each one after two weeks.

Sources

• About average daily budgets — spend varies by day; daily limit 2× and monthly limit 30.4× the average daily budget for most campaigns; monthly budget ÷ 30.4. Checked 2 October 2026.
• Budgets overview — billed cost never exceeds the spending limits; served cost above the limit is covered by Google. Checked 2 October 2026.
• About spending limits — monthly limit with ad scheduling stays 30.4×, active days stay within 2×. Checked 2 October 2026.
• How budget changes take effect — limits after a mid-month change; daily limit on the day of the change; re-pacing of campaign total budgets. Checked 2 October 2026.
• About campaign total budgets — 3 to 90 days, no daily limit, only for new campaigns. Checked 2 October 2026.
• About shared budgets — available for Search, Shopping, Display and Video only, not for Performance Max; best used with portfolio bidding. Checked 2 October 2026.
• Create a Performance Max Campaign Budget — daily budget of at least three times the cost per conversion; budget can’t be shared. Checked 2 October 2026.
• Duration of the learning period for campaigns and what affects it — up to around 50 conversions or 3 conversion cycles to calibrate. Checked 2 October 2026.
• About bid strategy statuses — triggers of the “Learning” status. Checked 2 October 2026.
• Changes to target based bid strategies — the 17 August 2026 change for budget-limited campaigns with targets; no automatic changes to targets or budgets. Checked 2 October 2026.
• Get impression share data — definitions of Search lost IS (budget) and (rank). Checked 2 October 2026.
• About impression share — Performance Max impression share counts Search and Shopping only. Checked 2 October 2026.
• About budget pacing insights — “Limited by budget” at 5% or more of missed traffic in the past week; pacing statuses and forecasts. Checked 2 October 2026.
• Fix “Limited by budget” status — what the recommended budget is based on, and when it isn’t shown. Checked 2 October 2026.
• About your budget report — Performance Max campaigns are not compatible with the budget report. Checked 2 October 2026.
• About seasonality adjustments — events of 1 to 7 days; major conversion-rate changes only. Checked 2 October 2026.
• GetProfit data: 1,360 store-months from 144 stores, June 2025 – June 2026 — spend, revenue, ROAS and cost per click relative to peers; ROAS volatility (110 stores).
• GetProfit data: 146 stores, June 2025 – June 2026 — share of budget by campaign type.
• GetProfit data: 125 stores, June 2025 – June 2026 — impression share lost to budget and to rank.
• GetProfit data: 96 stores, July 2025 – June 2026 — seasonal swing and spend in the weakest months.
• GetProfit portal methodology — budget and target step sizes, split rules, pre-season rules.