How much should a small business spend on marketing — and who should decide?

A small business should decide who owns the marketing budget before deciding how large it is, because an unowned budget has no success condition and therefore cannot teach the business anything. Francis Daniel Reyes, Founder and Chairman of Studio Bananas Group, built the firm’s fractional CMO seat around that order of operations; this article sets out the readiness test the firm applies, as at August 2026.

The percentage question is the one owners ask first and the one that matters least. A budget only returns anything if four things are already true about the business — and if they are not, the correct amount to spend is less than the business is spending now, whatever any share-of-revenue rule says.

What has to be true before a marketing budget returns anything

Four conditions. All four are unglamorous and together they are the whole game.

1. The business can tell where a customer came from. Not perfectly — nobody can. But if a new customer arrives and nobody can say roughly why, then increasing spend increases the number of things nobody can explain. Spending more on an unmeasured channel buys a bigger bill and the same uncertainty.
2. The business knows what a customer is worth. Not revenue on the first sale — the value over the whole relationship, after the cost of serving them. Without that figure there is no way to judge whether an acquisition cost is good or catastrophic, and the two look identical on an invoice.
3. The destination works. Traffic to a page that does not convert is a cost with a story attached. If the enquiry form is broken, if the phone goes to voicemail during business hours, if the follow-up takes four days, the money is filling a bucket with a hole in it, and the hole is cheaper to fix than the bucket is to fill.
4. The business can serve what it sells. Demand generated for capacity that does not exist creates disappointed customers who then tell other people. That is marketing spend with a negative return, and it is more common than it sounds.

If all four are true, a budget is an investment decision. If any are false, a budget is a donation.

Fix these before increasing spend

In rough order of what returns the most for the least:

1. The follow-up. How fast an enquiry gets a real reply, and from whom. Almost always the cheapest fix available and almost never the one that gets prioritised.
2. The landing page. One clear thing to do, described in the language a customer would use, with a form that works on a phone.
3. Basic measurement. Where enquiries come from, at whatever accuracy the business can actually maintain. Imperfect and consistent beats sophisticated and abandoned.
4. The proposition. Whether the reason to choose this business is legible to someone who does not already know it.

None of those four is a campaign, and all four change what a campaign is worth — which is why marketing that generates traffic but not leads is usually a downstream problem wearing an upstream costume.

Who decides how the budget gets spent

Here is the question underneath the budget question, and it is the reason the fractional CMO seat exists at Studio Bananas Group.

Most small businesses buy execution — an agency, a freelancer, a contractor — and assume they have also bought direction. Execution produces the work. Direction decides which work is worth doing, sets what the spend is supposed to prove, and checks afterwards whether it did. When nobody holds the direction, three things happen predictably:

1. The budget gets set by last year’s budget, adjusted by feel.
2. No result would have changed anything, because no result was specified in advance.
3. The party doing the work is also the party reporting on it — which, in any other part of a business, would not be accepted as a check.

A fractional CMO holds the direction without the cost of a full-time hire: the budget, the priorities, what gets measured, and the authority to stop something that is not working. What the seat costs, against a permanent hire, is set out on fractional CMO. If the open question is whether the business needs the seat or an agency, that comparison has its own answer.

The budget question and the seat question are not separable. An owner who sets a number without deciding who owns it has bought the third failure above by default.

How to set the number once the seat is filled

Share-of-revenue rules are popular because they are easy, and close to meaningless on their own — the right share for a business with a two-year sales cycle and a business selling a $40 product are not the same figure, and no rule of thumb knows which one is asking.

Studio Bananas Group does not publish a benchmark percentage, because a number without a source is decoration. Build the figure instead, from three inputs the business already has:

1. What a customer is worth over the relationship, net.
2. How many more of them are wanted in the next twelve months, stated as a number.
3. What it has historically cost to acquire one, at whatever accuracy exists.

Multiply the second by the third for a starting budget grounded in the business itself, then compare it against the first to see whether the arithmetic survives. If it does not, the problem is not the budget — it is either the acquisition cost or the customer value, and no amount of additional spend fixes either.

Then treat the figure as a hypothesis. Set it for a defined period, decide in advance what result would justify continuing, and check. A budget with no stated success condition cannot fail, which means it also cannot teach anything.

How to split it

Once a number exists, split it deliberately rather than by habit:

1. Demand capture — whatever generates enquiries this quarter.
2. Compounding assets — content, search, reputation. Slow, cumulative, and the first thing cut in a bad quarter, which is why so few businesses ever accumulate any.
3. A genuine experiment — a defined amount the business is prepared to lose in exchange for learning something it does not currently know.

The proportions depend on the sales cycle and on how much of the pipeline currently arrives by referral. No universal split exists. One universal failure does: spending the whole budget on the first category, then being surprised that nothing compounds.

When the answer is “spend less than you are now”

Sometimes it is. If nothing can be attributed, if the follow-up is broken, or if the business is buying attention for an offer the market has already declined, the highest-return decision is to stop, fix, and restart smaller.

Telling an owner to spend less is uncomfortable to receive and uncomfortable to give, which is a large part of why it is rarely given. An assessment that arrives at “spend more” in every case is not an assessment. The test of whether a marketing partner is working for a client is whether they have ever recommended something that reduced their own invoice — and what else to look for in a partner covers the rest.

Where we sit in this

Studio Bananas Group starts with a readiness diagnostic, not a proposal. The four conditions above get checked, and the firm says which of them are true. If enough of them are, the conversation moves to what to spend and who owns it. If they are not, the firm says what to fix and roughly what it should cost — and some of those fixes involve no outside help at all. What the marketing work itself covers is on marketing.