Set your Google Ads budget from three checks, in this order. Work back from the customers you want to the media spend that should produce them. Check that spend against the ceiling your margins can carry, then against the cash you can commit.
There is no universal starting spend.
Start with numbers you can explain
A budget built on numbers you can explain survives its first bad month. One built on a guess gets cut the first time the phone goes quiet. Before any arithmetic, gather the six inputs below. Record where each one came from and when.
Input | How to obtain or define it |
|---|---|
Monthly goal | New paying customers or genuine enquiries for a named service and period, within fulfilment capacity |
Average CPC | Recent relevant account data first, or a Keyword Planner forecast matched to area, network, dates and intent |
Click to enquiry rate | Genuine enquiries divided by clicks for the same traffic, with spam and duplicates removed consistently |
Enquiry to customer rate | From a matured enquiry cohort with the sales cycle delay stated, never this month's sales over new leads |
Contribution per customer | Revenue less the variable costs of delivering the work, before acquisition spending, on one GST basis |
Costs and available cash | Media, management, creative, tracking and setup separated, with operating cash and tax reserved first |
Research relevant click costs
When relevant account data exists, it beats any forecast. Without it, build a Keyword Planner plan for your keywords, location, network and dates. Then read its forecast Avg. CPC.
The top of page bid columns are historical bid ranges. Their midpoint is not a forecast of your average CPC. Our guide to Google Ads keyword research covers building the keyword list the forecast depends on.
Our guide to how much Google Ads costs in Australia covers market prices and what drives them. For a rough sense of scale, look at our Google Ads cost per lead by industry table. Its Keyword Planner ranges for Melbourne put standard plumbing at $8.14 to $51.85 a click. Family law sits at $14.64 to $54.57.
Most of its ranges start under $10 a click and top out between $10 and $55. None of these is a forecast for your keywords.
Define a genuine enquiry
Every rate in the budget depends on what you count as an enquiry. A click on a phone number is not an answered call. And a form entry from a bot is nothing at all.
Count a genuine enquiry as a new, relevant contact from a real prospect. Deduplicate and clean enquiries the same way every month. Our guide to a good cost per lead for Google Ads sets out the qualification definitions. It also explains how to judge what each lead should cost.
Across the same managed accounts, 20% to 30% of raw form entries are spam, duplicates or existing customers. That is why a budget built on raw entries is a budget built on sand.
The enquiry to customer rate needs a matured cohort. Take the enquiries from a period long enough ago that their sales decisions have been made. Count how many became customers, and state the delay. Most of this month's sales came from older leads. So dividing those sales by this month's new enquiries produces a rate that means nothing.
The estimator has two models that should not be confused. The quick industry and location model uses planning assumptions for cost per enquiry, not market averages. It excludes fees. It also shows a chosen scenario band rather than a measured range.
The section headed "Plan with your own acquisition numbers" is the one to use once the inputs above exist. It asks for:
customer goal
click to enquiry rate
enquiry to customer rate
low, base and high CPC assumptions
monthly fees and creative costs
It does not calculate gross profit, net profit, break even CPC or ROAS. The economics are explained here. Our return on ad spend tool does its own job.
Work back from your customer goal
One example service business runs through the rest of this guide. Its inputs are chosen for teaching, not taken from Elev8d results or industry norms. Figures are AUD excluding GST unless a cash total says otherwise.
The model assumes enough relevant demand and capacity exist. Real conversion rates and acquisition costs move as spend changes. The review section returns to this.
Assumption | Value and meaning |
|---|---|
Customer goal | 10 new paying customers a month |
Enquiry to customer rate | 25% of genuine enquiries become customers |
Click to enquiry rate | 8% of ad clicks become genuine enquiries |
Average CPC | $12 |
Contribution per customer | $1,000 before acquisition spending |
Other acquisition costs | $1,000 a month for management and tools in this simplified model |
Ten customers divided by 25% is 40 enquiries. Forty divided by 8% is 500 clicks. Five hundred clicks at $12 is $6,000 of media.
Run it forward to check. $6,000 buys 500 clicks at the assumed CPC. At the assumed rates, those clicks produce 40 enquiries and 10 customers. The ad only cost per enquiry is $150 and the ad only cost per customer is $600. Add the modelled $1,000 fee and the cost per customer is $700.
The spend in the diagram is media only. The contribution is before media and before the $1,000 fee. So the multiple the diagram shows is contribution against media spend, not return on investment.
A common mistake here is to read the Google Ads Cost/conv. column as the cost per customer. It is the cost per whatever action you set up, usually an enquiry. In this example that is $150, not $600.
Find your acquisition cost ceiling
Contribution is revenue less the variable cost of delivering the work, before acquisition spending. Ten customers at $1,000 contribution each bring $10,000. After $6,000 of media and $1,000 of fees, $3,000 remains for unallocated overheads, tax and profit. Do not call that $3,000 net profit. It is what is left before the costs the model ignores.
The ceiling works like this. At the goal volume, the most the media could cost before the campaign returns nothing is $9,000. That is $10,000 of contribution minus the $1,000 fee. Spread over 40 enquiries, the ceiling is $225 an enquiry. At an 8% enquiry rate, that is $18 a click.
Another way to the same place: the total acquisition ceiling per enquiry is $1,000 multiplied by 25%. That comes to $250. The fee takes $1,000 divided by 40, or $25 of it. That leaves $225 for media.
That is a zero surplus ceiling under the model, not a target. Set an operating target well below it. The gap leaves room for overheads, uncertainty and the profit the business exists to make. If the fee varies with spend, solve with the actual fee formula rather than treating it as fixed.
The chart holds the click to enquiry rate at 8% and moves the CPC between $8 and $18. At the assumed 25% close rate, the bars run from 5 to 15 expected customers. At 40 enquiries, the top of the bar is the $9,000 ceiling.
The ends of each bar are chosen CPC scenarios, not confidence intervals, benchmarks or recommendations. The values exclude fees and GST. The enquiry rate and the CPC rarely hold still when the budget doubles. So work the ceiling out again at each volume.
For an online store on Google Ads, the lead to customer step disappears. Required orders divided by the purchase conversion rate, multiplied by the CPC, gives the media spend. Contribution means revenue less product cost, fulfilment, transaction fees and expected returns. Our break even ROAS calculator does that sum once the margin is known.
Where repeat purchases or repeat jobs are part of the case, treat them as a cohort assumption. Date it, back it with evidence and give it a payback period. Then check when the cash arrives. Our lifetime value and payback calculator is built for that. A customer who is worth $3,000 over three years still has to be paid for this month.
Check what you can afford to commit
The goal says $6,000 of media. The bank account gets a vote too. Work backwards from the cash the business can commit after reserving operating cash and tax. Separate media, fees, setup and GST before comparing it with the plan.
Say the business can commit $5,000 excluding GST. Subtract the $1,000 fee and $4,000 of media remains. At unchanged assumptions that buys about 333 clicks. They produce 27 enquiries and 6.7 expected customers, so six or seven. The 10 customer plan needs $2,000 more.
There are three choices. Revise the goal to six or seven. Narrow the scope to the service with the best economics. Or improve the conversion rates before committing spend the business cannot carry. Expected fractions are averages across months, not a promise of two thirds of a customer.
Some businesses come to us with a budget already in mind. Their first plan is revised after the first mature cohort in about 1 in 3 cases. Usually that is because the close rate was a hope rather than a measurement.
GST is part of the cash and not part of the performance. Google's tax page states that all Google Ads sales in Australia are subject to 10% GST. This applies to accounts with an Australian business address. In the example, $6,000 of media plus $1,000 of fees is $7,000 excluding GST. That is $7,700 in cash if both attract 10% GST.
Setup costs are excluded from that example and go on top where they apply. The GST is not media available to Google. On the same assumption, the affordable $5,000 plan needs $5,500.
Check the actual supplier invoices, any one off setup and the GST treatment of each item. Do not apply the assumption to everything.
Whether the GST comes back depends on registration and eligibility. Both are set out in the ATO's conditions for claiming a GST credit.
A credit on the next activity statement does not remove the cash requirement this month. If GST cannot be recovered, it belongs in the economics consistently. Do not add it twice. And do not count a future lifetime value estimate as cash available today.
When a provider quotes a minimum, keep one distinction clear. A provider's commercial minimum is not a Google platform requirement. Elev8d recommends at least $3,000 a month for a managed account. A self managed test in a low CPC niche can start nearer $1,000 to $1,500.
The $3,000 is about what our Google Ads management service needs to be worth its fee. It is not about what Google needs. The fee itself is covered in our guide to Google Ads management pricing.
Affordability may land you well under those figures. Under narrow conditions, a few hundred dollars a month can still teach you something. Our guide to the minimum budget for Google Ads covers those conditions.
We'll check your tracking and click costs against the plan before you commit the cash.
Set daily budgets without surprises
Google Ads spends against an average daily budget. The word average does the work. Google's spending limits page says most campaigns have a daily spending limit of twice the average daily budget. The monthly spending limit is the average daily budget multiplied by 30.4. That is the average number of days in a month.
Served cost can exceed those limits. But you never pay more than them, because Google covers the difference. The limits are per campaign, not an account wide cap. They also do not guarantee the money is spent.
For the $6,000 media plan, $6,000 divided by 30.4 is about $197.37 a day. If the rounding has to stay under a strict $6,000 ceiling, it is $197.36.
Those are advertising charge limits for the average daily model. They are not the cash total with tax and fees. Nor do they apply to every campaign type, as the next paragraphs explain.
Budget edits change the arithmetic. Google's page on how budget changes take effect works out a revised monthly limit. It is the amount already spent, plus the new daily budget multiplied by the remaining calendar days. The day of the change counts. Google sets the daily limit on the change day from the highest average daily budget chosen that day.
Lowering a budget later in the day does not undo that exposure. For example, take $1,200 already spent plus $80 a day for 16 remaining days. That is a $2,480 limit for the month, not $80 multiplied by 30.4.
A reduced schedule no longer means a reduced allowance. Since 1 June 2026, Google paces campaigns that switch off whole days towards the full monthly limit. It does so regardless of the schedule. On a schedule that skips about a third of the month, the difference is large.
Until 1 June 2026, paced towards $2,000 for the month
From 1 June 2026, paces towards $3,040, the full monthly limit
The daily cap of twice the budget still applies on the days that run. Google's own example says to set about $66 a day to hold a 20 day campaign near $2,000. Schedules that turn off hours within a day are not affected. Do not divide a monthly target by business days and assume the schedule controls the spend.
Campaign total budgets work differently again. Google's campaign total budgets page describes a fixed amount for a campaign with a start and end date.
Campaign total budgets are available for new Performance Max, Search, Shopping, Demand Gen and YouTube campaigns. The budget is paced across the period, with no daily spending limit. Google promises you are never charged more than the total.
Existing campaigns cannot switch to a campaign total budget. And the budget type cannot be changed after creation. The 2 times and 30.4 times rules do not apply to this model. So do not mix the two sets of sums.
The payment threshold is the last confusion. In a postpay account, Google charges your payment method on the first day of each month. It also charges when accrued costs reach a threshold.
The threshold decides when the card is charged. It is not a spending cap. Bank charges rarely line up with the performance month. So check the payment setting, the invoice and the available cash together.
Split spend across services and seasons
Allocate the media budget to the demand the business can fulfil, in roughly this order of evidence. Services with proven demand, good contribution and a track record of genuine enquiries first. Then services where the business has capacity and a reasonable case. Then a bounded test.
Keep each allocation large enough to produce observations worth reading. Six campaigns sharing $6,000 will each spend too little to teach you anything. Separate brand and non brand results so that cheap branded enquiries do not conceal expensive acquisition elsewhere.
Resist any universal split between Search, Performance Max and remarketing. That's because the right split is the one the economics of each service support.
The example business puts $4,500 behind the service with the proven economics. It puts $1,000 behind a second service with capacity and a plausible case. And $500 goes into a defined test with a stop point. That is a decision, not a best practice percentage.
A month of mature results would change it. If the second service converts as well as the first, the split narrows. If the test produces nothing, it ends.
Use separate campaign budgets wherever a service needs a firm allocation. A shared budget moves unused funds between compatible campaigns and will not preserve the split. Google's shared budgets page limits them to Search, Shopping, Display and Video campaigns. Performance Max and campaign total budgets are excluded.
Our guide to Google Ads account structure covers when a service earns its own campaign.
Plan seasonal changes from these, not from a reflex:
recent monthly demand
search forecasts
booking lead times
stock
staffing
A Melbourne cooling business may need enquiries before the first hot week, but only while it can install. A quiet month does not by itself justify forcing spend into searches that are not happening.
Do not add 20% or 30% to every peak period by habit. Model the demand and the margins for that period. Reserve the cash. Then review the effect of each budget edit. A mid month edit changes the monthly limit the way the previous section described.
Review results before you scale
Define the review before the test starts, not when the first invoice arrives.
Media and cash ceiling: [media per month, fees, GST and the total cash committed]
Services and campaigns: [what is funded, with each allocation]
Assumption range: [CPC, click to enquiry rate, close rate and contribution, with sources and dates]
Tracking checks: [conversion actions verified, calls counted, spam removed]
Sales cycle lag: [how long before an enquiry cohort is mature]
Early warning checks: [tracking faults, irrelevant demand and cash risk, reviewed weekly]
Decision owner: [who can change the budget or stop the test]
Review date: [when mature outcomes are assessed]
Check early for the things that fail fast. A conversion action that stopped recording. Search terms that have nothing to do with the service. Spend running ahead of the cash plan.
Judge acquisition performance only once enough enquiry cohorts have matured. Allow 4 to 8 weeks for a first read and about 90 days to judge cost per lead. Allow longer when the sales cycle is long. Our guide to how long Google Ads takes to work explains this. A low volume test may stay inconclusive for longer than anyone wants.
Google's Target CPA guidance recommends evaluating that strategy over the last 30 days, with at least 30 conversions. It also says advertisers can start the strategy with no conversion history. Our rule of thumb is 15 to 30 conversions a month before trusting a target. Our guide to Smart Bidding strategies explains this. Google itself sets no minimum for Target CPA.
Compare the plan with what happened on the same definitions:
media spend
genuine and qualified enquiries
new customers
contribution
modelled acquisition cost
A 0% conversion rate can mean no results, a tracking fault or conversion delay. So investigate before declaring a cause.
Before launch and before any review that matters, set up and test your Google Ads conversion tracking. A budget decision made on numbers nobody has checked is a guess with a spreadsheet attached.
The reliability of that data has a privacy side too. Check whether the business is covered by the Privacy Act using the OAIC's small business guidance.
Configure tracking to collect only what the measurement needs. Keep the notices accurate too. The OAIC's tracking pixels guidance expects both. A privacy policy link on its own is not compliance.
The tree is a management aid, not Google policy. An inconclusive low volume test is not a proven failure. Increase spend when the economics, the lead handling, the cash and the available demand all support it. Judge the increase by the cost of the additional customers it buys rather than the blended average.
The next thousand clicks can cost more and convert worse than the first thousand.
If campaigns already underspend their budgets, raising the budget does little. The question is then demand, eligibility, targets or relevance. Reduce or pause when the agreed cash limit or the unit economics cannot support continuing.
The detailed method is in our guide to how to scale Google Ads. What if the ceiling and the cash never meet at any volume? Our guide to whether Google Ads is worth it for a small business has the honest answer.
Your monthly budget checklist
Approve a budget only when every row has an answer.
Goal
New customers or genuine enquiries for a named service, with fulfilment capacity confirmed
Dated inputs
CPC source and date, click to enquiry rate, matured close rate and contribution basis recorded
Acquisition ceiling
Media only ceiling per enquiry calculated and an operating target set below it
Cash including GST
Media, fees, setup and GST totalled against committed cash, with operating cash reserved
Budget type and daily figure
Average daily or campaign total chosen, daily figure set from 30.4, schedule effect understood
Allocation
Each service funded enough to learn from, brand separated, test bounded with a stop point
Tracking
Conversion actions verified, calls counted, spam and duplicates removed consistently
Review owner and date
Decision owner named, early checks scheduled and the mature review dated
If the account is already running, score it before you approve more spend. Money added to a leaking account leaks faster. Our free scorecard checks tracking, structure, keywords, ads and optimisation. It treats verified conversion tracking as a prerequisite.
What we recommend at Elev8d
Start from the customer goal and the margin, never from a round number. We run the three checks in the same order every time: goal, ceiling, cash. When the cash can't fund the goal, we shrink the goal or narrow the services. We don't nudge the close rate up to make the sum work.
Measurement comes before money. If calls matter, we set up call tracking for Google Ads and clean the enquiry count. We do both before the first dollar of media goes out.
Then we fund the strongest service properly and leave the rest until it has mature results behind it. One campaign that can learn beats six that each starve. We set the daily figure from 30.4 and leave it alone between reviews. That's because every mid month edit changes the monthly limit.
We take a first read at 4 to 8 weeks and judge cost per lead around day 90. Only then do we raise the budget, in steps. We watch what the extra customers cost rather than the blended average.
Below $3,000 a month of media, we'd point you to a self managed test, not a management fee. Run the test in a narrow, low CPC niche. The ceiling and the cash may never meet at any volume. If so, we'll tell you Google Ads isn't the channel for now.
Next steps
Write down the six inputs with their sources
Pull your own recent CPCs or a Keyword Planner forecast. Take the close rate from a matured cohort. Record the contribution per customer and the date of each input before touching a budget figure.
Run the inputs through the estimator
Enter the inputs in the estimator's own numbers section. It shows the low, base and high CPC cases side by side. Then settle the cash figure with fees and GST included.
Book the review before the campaign goes live
Write the test plan with its decision owner. Then book the mature review before the first click is bought.
Sources
- False or misleading claimsaccc.gov.au
Australian Competition and Consumer Commission. Claims about future matters, such as forecasts and projections, need reasonable grounds.
