The fastest way to plan a marketing budget is to start from the sales you need next year and work backwards to what that many customers costs to buy. Starting from last year's number, or from a percentage a survey once printed, gives you a figure with no connection to your own margins. I plan paid media budgets for businesses that hold real money and want it to earn something, and the plan below is the order I actually work in.
Work backwards from the sales you need
A marketing budget is the cost of acquiring a number of customers, so the plan starts with that number and ends with money. The chain is: customers needed, leads needed to produce them, clicks or impressions needed to produce the leads, and finally the price of those clicks.
Here is the arithmetic with made-up round numbers, to show the method and not to set a benchmark. Say a business needs 20 new customers a month. If one in four qualified leads becomes a customer, it needs 80 leads. If one in ten visits to the landing page becomes a lead, that is 800 visits. At an average click price of $5 the media line is $4,000 a month. Change any of the four inputs and the budget moves, which is exactly why it beats a flat percentage.
The weakest input is nearly always the lead-to-customer rate, because marketing rarely sees it. If you do not know it, the first job of the budget is to find out, and the CRM connection between ads and closed deals is usually where that number comes from.
Decide what a customer is allowed to cost
The ceiling on your budget is the most you can pay to acquire one customer and still be happy with the result. Work it out from gross profit per customer over the period you are willing to wait, not from revenue, and keep some room for everything else the business has to pay for.
I go through this ceiling in more detail in the piece on customer acquisition cost against lifetime value. For budgeting, what matters is that the ceiling exists before the first dollar is allocated. A plan that says "we will spend $10,000 a month" without a per-customer ceiling cannot tell you in month three whether to spend more or stop.
Keep media, people and fees on separate lines
A marketing budget contains at least three different kinds of money, and mixing them is how businesses lose the ability to say whether a bad month was a market problem or an execution problem.
- Media spend: what you pay Google, Meta, LinkedIn or a publisher for clicks and impressions. This is the line that scales with results.
- People and tools: salaries, freelancers, software licences, call tracking and analytics. It stays roughly flat whether media goes up or down.
- Management fees: what an agency or consultant charges to run the media. I quote a fixed-price audit or a month-to-month retainer, so the fee does not grow just because spend does.
When a vendor quotes a single all-in monthly number, ask for the split. If the fee is a percentage of media spend, the vendor earns more when you spend more, which is not the incentive you want during a month when the honest advice is to pull back. The comparison of hourly, retainer and percentage pricing covers how each model behaves.
What a percent-of-revenue rule is good for
A percentage of revenue is a sanity check on the finished plan, not a way to build it. Gartner publishes an annual CMO Spend Survey that reports company marketing budgets as a share of revenue, and the share moves from year to year and by sector, which already tells you it is a description of what other companies did and not a target for you.
Use it in one direction only. Build the plan from sales, then compare the total to the published range. If yours is far above it, you are either in a growth phase or you have not tested your lead-to-customer assumptions. If it is far below, you may be assuming a conversion rate you have not earned yet. Either way, the comparison prompts a question and does not answer one.
Splitting the money: proven, testing and held back
The 70/20/10 rule that people search for says to put 70 percent into channels that already work, 20 percent into promising ones and 10 percent into experiments. It is a reasonable default for a business that already has a working channel, and it fails for one that does not, because there is no proven 70 percent to protect.
What I do instead is split by evidence:
- Proven: campaigns where cost per customer is already inside the ceiling. They get funded first and scaled in steps.
- Testing: a new channel, audience or offer, with a fixed test budget and a written pass mark before it starts.
- Held back: money left unallocated at the start of the year. When a test works, or a competitor leaves an auction, you can act in the same month rather than waiting for the next planning cycle.
The "3-3-3 rule for marketing" also shows up in searches, but it is used to mean different things by different authors, so I do not plan with it and would not trust a plan that leaned on it.
Do not let brand search flatter the plan
Paid search reports blend two very different kinds of click: people searching for your company by name, and people who have never heard of you. In the accounts I run, brand terms often take 20 to 35 percent of clicks, and they convert far better than everything else. Blended together they make the whole account look cheaper than acquisition really is.
For budgeting, that means planning brand and new-customer spend on separate lines, with separate targets. Brand spend is defensive and small. The money that grows the business is the new-customer line, and it is the one your cost-per-customer ceiling has to be tested against. The write-up on cross-channel attribution explains why the same sale can be claimed by three platforms at once.
What people actually search when they plan a budget
Search demand shows what owners want help with, and it is not what a strategist would guess. The table is US monthly search volume and keyword difficulty for this cluster, pulled from DataForSEO in September 2026.
| Search term | Monthly US searches | Keyword difficulty |
|---|---|---|
| marketing budget template | 390 | 11 |
| planning a marketing budget | 390 | 34 |
| marketing budget planning | 390 | 29 |
| marketing budget for small business | 140 | 28 |
| marketing budget percentage of revenue | 50 | 0 |
| how much should a marketing budget be | 40 | 35 |
Owners look for a template roughly eight times more often than for a percentage of revenue. They want a structure to fill in before they want a benchmark, which is why this article gives you the structure. Advertisers also pay a lot to reach this audience: DataForSEO lists an average click price near $30 for "planning a marketing budget" and "marketing budget planning", a sign that software vendors and agencies see buyers here.
Review the plan on a schedule, and change it for a reason
A plan is only useful if it changes when the evidence changes. I put reviews on the calendar: a short check every month, and a full re-plan each quarter.
The monthly check answers three questions. Is cost per customer inside the ceiling? Did the tests hit their pass marks? Is there a reason to release some of the held-back money? Move budget between lines only after enough time has passed for the leads to close, which for a business with a long sales cycle is longer than one month. Cutting a channel because last week looked bad is a common way to kill something that was about to work.
A small budget changes the arithmetic but not the method. The piece on whether $10 a day is enough for Google Ads shows what a very small budget can and cannot learn, and the guide to PPC management for small businesses covers who should run it. If you would like a second pair of eyes on the numbers behind your own plan, this is the kind of work I do through my Google Ads management service. Google's own Google Ads Help Center documents how campaign budgets are spent and paced day to day, which is worth reading before you set daily figures.
