How Much Should a Small Business Spend on Marketing?

An agency sends a quote, or an ad platform asks for a daily budget, and you realize you have never actually decided what marketing is allowed to cost. Until now you have spent when it felt necessary: after a slow month, before a launch, when a competitor showed up everywhere. Somewhere along the way you heard that a business should spend 7 to 8 percent of revenue, and you have no idea whether that fits a business like yours. Spend too little and it may never add up to anything. Spend too much and it can hurt the business that is paying for it. The question of how much a small business should spend on marketing can be answered with your own numbers, and it often takes about an hour. You may find the right figure is larger than you feared, smaller than you were quoted, or zero for now, and each of those is a real answer you can act on.

How much should a small business spend on marketing?

Set your budget from the customers you need and what you can safely pay to win each one, then use a percentage of revenue only as a rough cross-check. Start with one measurable goal, count every cost, cap a small test at an amount you can afford to lose, and raise spending only when sales and gross margin support it.

That answer rests on three numbers you can work out yourself: what a customer is worth to your business, what one currently costs to win, and the most you can lose on a first test without hurting the business. Once you have those, the budget stops being a guess. It becomes the number of customers you want, multiplied by a cost per customer you have decided you can afford, limited by cash you can spare.

No U.S. regulator or standards body publishes a percentage every small business should spend. The closest government guidance is the FDIC's Money Smart module on developing a marketing plan, which notes that budgets are often set arbitrarily as a share of revenue or last year's spending, and recommends goals, a cost for each line item, and revision once results come in. The sections below follow that same logic, one number at a time.

What is one new customer actually worth to your business?

The sale price is not the number to use. A $1,000 job at a 20 percent margin leaves $200 after the cost of doing the work, and that $200 is what you have available to pay for winning the customer. Spending $1,000 to win that customer would lose money even though the invoice looked big.

The more useful figure is contribution: what a customer leaves behind after the direct cost of serving them. A simple version works like this. Take your average revenue per purchase, multiply it by how many times a customer buys in a period, multiply that by how many periods they usually stay, then multiply the result by your gross margin. The American Marketing Association's lifetime value calculator separates exactly these two versions, revenue lifetime value and contribution lifetime value, and it also calculates payback: your acquisition cost divided by the contribution a customer produces each period. That tells you how long it takes to earn back what you spent to win them.

If you do not have repeat-purchase records you trust, use the first sale only. A $500 average first sale at a 40 percent gross margin gives $200 of first-sale contribution. That is arithmetic from your own figures, not a benchmark, and it is a cautious starting point because it ignores repeat business you have not yet proved.

What does it cost you to win a customer right now?

Customer acquisition cost, often shortened to CAC, is everything you spent to win new customers in a period, divided by the number of new paying customers you won in that same period. Two parts of that sentence trip people up.

First, count everything. Media spend is only part of it. Agency or contractor fees, creative work, the software you pay for to run campaigns, and relevant staff time all belong in the total. Pick one definition and use it every time, so this month's number can be compared with last month's.

Second, count customers, not leads. A click is not a customer, and neither is a form submission. If you divide your spend by inquiries, the cost will look far better than it really is. Divide by people who actually paid.

SCORE's marketing plan guide includes a worksheet built on exactly this. You divide your promotional cost by the new customers it produced to get a cost per new customer. Then you multiply that cost by the number of new customers you want, and the result is a budget tied to a sales goal. When your goal is sales rather than general awareness, it is the most directly usable method available.

To get the inputs, pull the last three to twelve months of marketing costs and sales records. Count the genuinely new paying customers and, where you can, note where each one came from. If you cannot tell where customers came from, that gap is itself the first thing to fix, and it matters more than the size of your budget.

How much can you afford to lose on a first test?

This is the number that protects you. Before choosing any channel, decide on an amount of cash the business can lose completely without affecting payroll, inventory, taxes or debt service. If the test produces nothing, you should be disappointed, not in trouble.

Then limit it a second time from the other direction. Take the most you are willing to pay for one customer and multiply it by the number of customers the test is meant to win. Using the example above, a cautious test might cap the cost of a customer at or below $200 until repeat buying is proven. Your test budget is the smaller of the two figures: what you can afford to lose, or what that customer goal justifies.

Watch the timing as well as the total. A budget can look affordable over a year and still be impossible in the month the bills arrive. Spread the amount into weekly or daily caps so the spending matches your cash flow instead of landing all at once.

How do you turn those numbers into a monthly budget?

With a customer's worth, your current cost per customer and a loss cap in hand, a monthly figure comes from putting them in order. The documented methods each answer a slightly different question, so it helps to see them side by side. After that, three rules decide where the money actually goes. They keep a small budget from being spread so thin that it teaches you nothing.

Fund what already works first

If one channel already brings in paying customers at a cost below your maximum, that is where the next dollar goes. It has evidence behind it that a new channel does not. Raise it in steps, check the cost per customer after each step, and stop raising when that cost climbs toward what a customer is worth or when you run out of capacity to serve new work.

Add one channel at a time

A new channel should enter as a test with one customer segment, one offer, and one clear definition of success, such as a paid booking rather than a click. Running several new things at once splits a limited budget so thin that none of them produces enough results to judge. Pause or defer everything but one or two until you know what the first one does.

Keep a floor that never lapses

Some costs sit underneath every budget decision: your domain, your website, and the accounts customers use to find and reach you. Letting those lapse to save money can cost far more to recover than it saved. If you ever have to cut back, the order matters: protect what is slow to get back, such as your domain, your email and your profile access, before you cancel anything.

What if your business has no sales history yet?

A new business cannot base its first budget on last year's revenue, because there is no reliable last year. SCORE's guidance for startups is to work from initial sales goals and revise the figures once actual sales exist. In practice, a first test looks like this:

• One segment: The single group of customers you most expect to buy.
• One offer: A specific thing at a specific price, not your whole catalog.
• One channel: The place those customers are most likely to see you.
• One conversion: A defined result, such as a paid booking, not a click or an inquiry.
• A fixed amount: Your loss cap, split into weekly or daily limits.
• A review date: When you will record spend, qualified leads, sales, revenue and gross margin, and decide.

Estimate what a first customer contributes and set a maximum test cost per customer below it, unless you have good reason to expect repeat purchases that justify a longer payback. Do not stop or change course on a feeling. Change it after you have written down what the money bought.

If paid search is the channel, know how the budget behaves. Google explains in its help page on spending limits that for most campaigns, a day's spending can reach twice your average daily budget, while the monthly limit is 30.4 times that daily budget. A $10 daily budget does not guarantee a $10 day, though the month stays bounded. If you want a sense of what clicks cost in the first place, our breakdown of how much Google Ads cost covers it. Set up conversion measurement in Google Ads before the test starts, so the sale you defined is the thing being counted.

Where do the 7 to 8 percent and 7.7 percent figures come from?

Once you have your own numbers, a percentage of revenue is useful as a cross-check. If your plan works out to far more than any published figure, look again at your assumptions. If it comes out lower, that may be fine. The two figures you are most likely to have heard come from very different sources, which is why they rarely agree with each other or with the ranges you find elsewhere.

The 7 to 8 percent rule, as it is usually repeated, also said the money should be split between building the brand and promotion. Because the SBA page it came from can no longer be found in the SBA's library, treat it as a historical rule of thumb with no stated research behind it, not as current federal guidance.

The Gartner CMO survey found budgets fell to 7.7 percent of company revenue in 2024, down from 9.1 percent in 2023. That is a solid measurement of what large companies spend. It says little about what a business with a few employees should spend, because the companies surveyed have very different costs, margins and goals.

If you come across another range, ask where it came from and who was in the sample. Then put it next to your own three numbers and let your numbers decide.

When is spending less, or nothing yet, the right call?

Sometimes the right amount to spend on ads this month is zero, and that is a sound decision, not a failure. Buying media makes sense only when the money can actually be turned into customers you can see. Hold off, or spend on the foundation first, if any of these are true:

• You cannot fund a meaningful test: The amount you can safely lose is too small to learn anything from the channel you want.
• You cannot define a sale or your margin: Without those, there is no way to set a maximum cost per customer.
• You cannot tell where customers come from: A test you cannot measure only tells you that money left.
• Your website or lead handling is broken: Forms that do not send, a checkout that fails, or inquiries nobody answers will waste whatever traffic you pay for.
• You cannot handle more work: More demand than you can serve can hurt the business you already have.
• The numbers do not fit the channel: If your maximum cost per customer is lower than what it likely costs to reach buyers there, that channel may not suit you right now.

In those cases, a smaller step usually fits better: a one-time fix to the website, a basic tracking setup, a referral process, or conversion rate work when visitors arrive but do not buy. Once the foundation holds, the three numbers will tell you when to spend.

A small budget is not pointless if it is narrow enough to answer one specific question, such as whether one offer can produce a qualified inquiry below a cost you set in advance. And a budget that matches some published percentage is not automatically enough. What matters is how what a customer contributes, what one costs to win, when the cash goes out, and whether you can measure the result fit together.

What can you finish this week to set your own number?

You can get the three numbers in about an hour if your records are in reasonable shape. The fixes around them take longer, and it helps to know roughly how long before you start. These are practical working estimates, not surveyed durations.

At the end, you should be able to say, in dollars, what you will spend, what result it is meant to produce, the most you will pay for a new customer, what the spend includes, and the date you will look at it again. That is a budget you can defend to a partner or a bookkeeper, and one you can pause without guessing.

Frequently Asked Questions About how much should a small business spend on marketing

What percentage of revenue should a small business spend on marketing?

No single percentage fits every business. Start from what a customer contributes, what one costs to win and the cash you can spare, then use a published percentage only as a rough check, with its source and sample in mind.

Does my marketing budget include agency fees?

Yes, if you want a true cost per customer. Include every cost of winning or keeping customers, such as media, agency or contractor fees, creative work and software, and use the same definition every month.

How do I calculate customer acquisition cost?

Add up what you spent to win new customers in a period and divide it by the number of new paying customers from that same period. Divide by customers who paid, not by leads.

What should my marketing budget per month be?

Take the number of new customers you want, multiply by the most you can pay to win one, and limit the result to what you can afford to lose. Spread that across weekly or daily caps so it matches your cash flow.

How long should I test a new marketing channel?

Until you reach the spend or decision point you agreed on before starting, while recording spend, qualified leads, sales and gross margin. One month's results alone rarely prove a channel works or fails.

Can a new business start with a very small marketing budget?

Yes, if the test is narrow enough to produce a useful answer without putting essential cash at risk. One segment, one offer, one channel and one defined sale can teach you more than a larger budget spread across many things.

The Bottom Line

A marketing budget does not have to come from a rule you cannot trace. It can come from three numbers: what a customer leaves behind after costs, what it currently takes to win one, and the most you can lose on a test without hurting the business. Put money behind what already works, add new channels one at a time, and protect the basics that must never lapse. The published percentages can check your math, but they describe other businesses, so they should not replace it.

Once you work this way, spending stops swinging between impulse and panic. You will know what each dollar is meant to buy, when to raise it, and when to stop, and you can say no to a quote with numbers behind the answer.

If you want a second set of eyes on those numbers, Web Leveling can help. Our digital marketing work starts with one written plan that says which channels, why each one, what it costs and what it is supposed to cause, and if a channel does not fit your business, we say so and skip it. Reporting is tied to qualified leads, sales and customer value, with every account in your name. We work with small and medium businesses across the country and overseas. Send us your budget question, and we will help you work out a number that fits.

Terms