Once the number is set, allocating a marketing budget comes down to a single question: what job does your business need done right now? Match the money to that job, whether that’s reaching a market that doesn’t know you yet, converting interest you already have, or keeping and growing the customers you’ve won. Industry benchmarks are great for reference, but the right split depends on your specific goals.
For a small business, deciding how much to spend on marketing is a hard call, and it isn’t the one I’m making here. There’s a rough benchmark to anchor to, and the institutional numbers land in a narrow band.
Different businesses sit at different stages, and a move that’s right for one is premature for another. A company that has run marketing-led growth for five years and one just getting its first real budget off the ground are solving different problems.
That stage lens runs through everything below.
Fund channels by their job
Decide what each channel is for before you decide how much it gets. Channel benchmarks describe what other companies did, not what your business needs, and running with benchmarks risks skipping the only question that matters: What job are you hiring this channel to do?
Most channels do one of a few jobs. Some get you in front of people who aren’t looking for you yet, like a LinkedIn post or a conference talk that reaches someone who didn’t know your kind of help existed. Some catch people who are already looking, like a Google ad or a search result that shows up the moment someone types in the problem you solve. Some turn warm interest into a decision, like a strong sales page or a quick follow-up after someone raises their hand. And some keep the customers you’ve already won. A channel that’s excellent at one of these jobs can be useless at another.
Once you see channels as jobs, where to put the money gets clearer.
If people are already searching for the problem you solve and just haven’t heard of you, that’s a strong reason to start with search. The intent is already there, and a Google ad or a page that ranks catches it. The only search spend with nothing to catch this early is bidding on your own name, since nobody is typing it yet.
The one case where the room really is empty is narrower: a product so new that people don’t yet know to search for it. There you have to get on their radar before there’s any search demand to capture, and that’s rarer than it feels, so be honest about whether it’s you.
If interest is coming in steadily but stalling before it turns into customers, that’s the opposite problem, and spending more to bring in new interest just makes the leak bigger.
Your existing stack shapes this too. If you’re not sure what your current tools are doing for each of these jobs, a martech audit is a good way to find the overlaps and gaps before you add spend.
Favor spend you can see working
One split is worth making early: spend that shows you whether it’s working, and spend that doesn’t. Both can pay off. What separates them is whether you get a signal soon enough to steer the money while it’s still in motion.
Some spend gives you a signal within weeks. A Google ad shows you clicks, form fills, and what each lead cost almost as soon as it runs. A new piece of content shows you impressions and engagement long before it produces a single customer, so you can watch it gain traction, or fail to, while there’s still time to adjust. In each case you get a read early enough to act on.
Other spend gives you almost nothing to read early. Put money into a radio spot, a billboard, or a podcast ad with no promo code, and for weeks there’s no engagement number to check and no traffic bump to point to. It might be working. Someone could show up a year later saying they heard you on the radio, which is a great outcome! But if that mention is the only trace they ever leave, you spent the whole year with no way to tell whether the spend was doing anything.
The practical move is to favor the spend you can read early, because that’s the spend you can steer. A Google ad tells you fast whether it earned its place. Ad space at a big industry event might be building real awareness or none at all. The issue is that the signal comes back slowly and second-hand, so you can’t easily do more of what’s working or stop what isn’t while the money is going out.
None of the slower-to-read spend is off-limits. What matters is the order you fund things in. Put the spend you can read first, and let the more experimental, feel-good bets come out of what’s left after that, especially if cash is tight and you need growth.
Importantly, a lot of the fuzzy stuff can be made less fuzzy, and it’s worth rigging up a way to measure it before you write the check.
Any of these make the “what’s working?” question a little less muddy.
If you’ve only recently put real budget into marketing, there hasn’t been time to see clear patterns yet, so don’t force a precision that isn’t there. Early on, expect the numbers to be thin. Be clear about what each dollar is meant to do, and lean toward the spend you’ll be able to track once volume builds.
How you’d measure any of this once you have the volume is its own topic, and a big one, so I’m saving it for a separate piece. But for spending the budget you’ve set, sorting your spend by what you can quickly see working is the first cut worth making.
Do the research before you spend
So far this has assumed you know your options and mostly have to choose among them. If you’re reading this thinking you don’t know the options, that’s fine, and it’s the more common place to start. You don’t have to guess your way through it. A little research up front tells you where the money is likely to work, and none of it requires you to already be a marketer. Here is where to look.
Start with people. Talk to other business owners who have spent on the channel you’re weighing, and ask plainly what it cost them and what it returned. Peers who don’t compete with you will often share what a benchmark never would. It’s worth the time to spend a few hours texting or emailing other operators and asking around.
Then study your competitors, who mostly won’t tell you anything, so you read it off what they do. Look at what they sponsor, the research or reports they publish, the webinars they run, and the content they put out week after week. Something a competitor keeps investing in is usually working for them, and when you see the same move across several of them, that’s a strong signal for where the money in your market tends to go.
Then use AI to fill in what your network can’t. A general assistant is good at the research a busy owner has no time for: the common ways to reach your kind of customer, what they tend to cost, and how long they take to pay off. The whole trick is asking specific questions, so here are a few you can adapt.
Before any money goes out, decide what you are looking for and how long you will wait to see it. Pick the signal that will tell you it is working, whether that is leads, booked calls, or closed deals, and pick the point at which you will judge it. Deciding this in advance is what keeps you from staring at the spend three months in and rating it on a feeling. It also shows you, up front, whether you have the patience and the runway the channel actually needs.
Then be patient, and get specific about what patient means. A channel that takes six months to show results will look like a failure at week three if week three was your expectation. The same research that told you what to try will tell you how long to wait: ask the people you talked to how long it took them, and have the AI give you typical ramp times. That way you set the waiting period in advance, instead of running out of nerve halfway through and pulling the money right before it would have worked.
Grow your existing accounts
Keeping customers is table stakes, and you’re presumably already on it. What’s easy to leave out of a marketing budget is the money you could put toward growing the customers you’ve already won. Selling more to an existing account, through more seats, a wider scope of work, or an adjacent service the same buyer needs, is often the fastest revenue you can get. There’s no one to win over from scratch, because they already trust you and see the value, so the path to yes is short.
When both are true, growing an account can move revenue faster than landing a new one. When satisfaction is shaky, or you’ve already sold them everything you do, pushing it there costs more in goodwill than it returns. Like everything else here, it’s stage-dependent: early on, with few accounts to grow, acquisition still leads, and once you have happy customers worth expanding, this is usually the cheapest growth you can buy.
When more budget isn’t the fix
Sometimes the constraint isn’t the size of the budget at all. Before adding spend, it’s worth checking whether money is the thing holding growth back, because pouring budget into a system that’s leaking somewhere else just produces more expensive leaks.
The usual culprits sit on either side of the spend. One is positioning: if people can’t quickly tell what you do and who it’s for, more traffic just means more confused visitors. The other is what happens after someone shows interest: if leads arrive but stall, or go unworked, more budget just buys more stalled leads.
When the issue is one of these, the fix is rarely a bigger number. It’s often a person or a process. Sometimes that means tightening how sales and marketing hand off to each other, which is more of an operations question than a spend one, and I’ve written separately about how those ops functions divide up.
Point the budget at the right problem
Setting the number is hard, but it isn’t what decides whether the money works. That comes down to whether the budget is pointed at the right thing. For some of you that’s getting in front of people who don’t know you yet, for some it’s converting interest you already have, for some it’s keeping and expanding the accounts you’ve already won, and for some it’s proving a single way of winning customers works before optimizing anything at all.
Regardless of your situation, deciding where to route your budget is tricky. These tips should help you feel a bit more confident as you start the process.