The Right Marketing Budget Is Not a Percentage

Ask the internet how much to spend on marketing and you will receive a ratio: five percent of revenue to maintain, ten to grow, more if you are a startup, with footnotes citing surveys of what other companies spend. The advice is comfortable, citable, and reasoned from nowhere. Spending what the average company spends produces, at best, average outcomes, and it produces them by accident, because the ratio contains no information about the only things that matter: what a customer is worth to you, what one costs to acquire, and how fast you are trying to grow. Twenty years of budget meetings have convinced me the percentage rule survives because it ends arguments, not because it wins markets.

Watch it fail in both directions with the same arithmetic. Company A does $10 million in revenue at 80 percent gross margin with a $900 fully loaded CAC against a $4,000 customer LTV; the folklore prescribes $700,000 or so of marketing. But every $900 this company spends returns better than 4-to-1 over the customer's life, and its market has depth; capping spend at seven percent is not prudence, it is a decision to gift growth to whichever competitor read their own unit economics instead of the survey. Company B also does $10 million, at 25 percent margins, with a CAC near its first-year contribution; the same seven percent prescription funds acquisition that destroys cash with every cohort. Identical revenue, identical percentage, opposite correct answers. The ratio was never the variable. It was the residue of variables nobody wrote down.

Here is the sequence that replaces it, and it runs bottom-up. Start with the growth target as a customer count, and hold the whole exercise loosely enough to reforecast, because the Q2 plan everyone signed off on in March is already wrong applies to budgets doubly. Take the: revenue goal, minus what retention and expansion of the existing base will produce, divided by revenue per new customer, equals customers marketing and sales must manufacture. Multiply by your fully loaded CAC, computed with the honesty I argued for in CAC is not CPA, and you have the acquisition budget the target actually requires. Add the spending that CAC math does not capture but the future depends on, brand and demand creation, the relay leg I described in Google Ads vs. Meta Ads is not a rivalry, typically 20 to 40 percent on top of pure acquisition for companies playing beyond the current quarter. Now compare the total to what the P&L can carry. If they fit, you have a budget derived from reality, one you can defend line by line, which is the entire thesis of stop presenting media plans before you know the margin. If they do not, you have discovered something the percentage rule would have hidden for another year: either the growth target is fantasy at current unit economics, or the unit economics are the actual project, and no budget ratio fixes a CAC-to-LTV problem.

Here is the sequence on real numbers. A client targeting $4 million in new revenue, at $8,000 average first-year value, needed 500 new customers; retention and expansion covered nothing, they were early. Fully loaded CAC ran $1,400, so acquisition required $700,000; we added 30 percent for demand creation, $210,000, for a bottom-up budget of $910,000 against roughly $9 million of expected total revenue. Ten percent, as it happened, coincidentally near the folklore, but derived, defensible, and, most usefully, falsifiable: when Q2's marginal CAC data came in under plan, the same arithmetic justified accelerating to $1.1 million while competitors held their ratios and their market share.

Two constraints modify the output, and both are about ceilings the spreadsheet cannot see. Markets saturate: your CAC is a curve, not a constant, and each increment of spend buys customers slightly worse than the last, so budgets should be sized to the point where marginal CAC approaches your allowable, not to the point where the average still looks fine, the margin-versus-blend trap from what is a good ROAS wearing a budget costume. And organizations saturate: a team that can deploy $200,000 a month well cannot necessarily deploy $500,000 well next month; creative pipelines, landing page capacity, and sales follow-up all have throughput limits, and money past the limit converts to waste at a remarkable rate. The best budget expansions I have run were staircases with measurement landings, each step held until incrementality testing, the kind any team can run per you do not need a data science team to run a holdout, confirmed the marginal dollar still cleared the bar.

What remains useful about the percentage? One thing: as a sanity flag after the real math is done. If your bottom-up budget lands at one percent of revenue, you are probably harvesting a brand someone else built and calling it efficiency; if it lands at forty percent, you had better be a venture-backed land grab with the retention curves to justify it. Between those tripwires, the ratio has nothing to teach. Budgets are not a percentage of the past. They are a purchase order for a specific future, and the correct amount is whatever that future costs at your unit economics, staircased to your team's ability to spend it well.