The 11 most common questions in food & beverage marketing.

The percentage-of-revenue rule is the most repeated answer in marketing and the least useful one in food and beverage. It assumes your constraint is money. It usually isn't. Your constraint is signal — whether the platform gets enough conversion events, on enough distinct creative, to figure out who your buyer is.

Below that threshold, more budget doesn't buy performance. It buys a longer, more expensive learning phase on creative that was never going to work. We've watched brands triple spend against a single hero video and get a worse blended CPA than they started with.

So the question to ask isn't "what percentage should we spend." It's three sharper ones: How many distinct creative concepts can we produce and refresh each month? What's our contribution margin per unit, and how many units does one new buyer need to purchase before we're whole? And how long can we fund the gap between acquisition and payback?

Answer those and the budget number falls out of the math. Skip them and you're guessing with a spreadsheet formula that was written for a category with none of your economics.

A 34-year-old woman in Austin with a household income over $100k might buy your product for health, for convenience, for her kids, for taste, or because it signals something about her. Five motivations, one demographic. An ad built for the demographic speaks to none of them clearly, so it speaks to all of them weakly.

This is what the Why People Buy Pyramid exists to sort out. Every ad leads with a motivational tier — basic function, emotional payoff, personal growth, or something beyond the self. When we classified the corpus by that tier, the distribution was lopsided in a way that explains a lot of underperformance:

Nearly three in ten F&B ads lead with basic function — what the product is and what it does. That's the easiest brief to write and the easiest one for a competitor to match. Meanwhile the tiers that create preference rather than comparison are where the field thins out.

None of which means "go emotional." It means the tier should be chosen because research says that's what drives your buyer, not because it's what the category defaults to.

Brands consistently overestimate how long an asset lasts, usually because the asset took six weeks and three approval rounds to produce. The corpus is unsentimental about it:

Two things follow. First, your production system matters more than any individual asset — a brand that can ship twelve concepts a month will beat a brand that ships one brilliant one, because eleven of yours are still alive when theirs dies. Second, the 3.4% of ads that run past six months are worth studying closely. They're almost never the polished brand film. They're usually a specific, unglamorous message that keeps working because it's answering a real motivation rather than decorating one.

Refresh cadence should be a calendar commitment, not a reaction. By the time frequency has climbed and CPA has visibly degraded, you've already paid for the fatigue.

A $6 grocery SKU and a $60 DTC subscription box are not the same advertising problem, and they don't belong on the same channel mix. The $6 SKU can't absorb a $40 acquisition cost, which rules out most of what works for the subscription. The subscription can afford to buy attention outright, which the SKU cannot.

Two questions settle most of it. What can one new buyer be worth in the first 90 days? And how often does this product get repurchased without being re-sold? High value plus high repeat can justify almost any channel. Low value plus low repeat needs channels where the creative does the persuading and the media is cheap enough to lose on the first purchase.

The failure mode we see most isn't picking the wrong platform. It's picking five — spreading budget so thin that none of them ever exit the learning phase, then concluding that all five "didn't work." Concentration beats coverage until you have proof.

Here's how the corpus breaks down by creative approach:

Two structural notes from the same dataset. Copy runs short — 48.7% of ads keep primary text between 50 and 150 characters, and only 4.4% exceed 600. And the call to action is nearly uniform: 59.6% of F&B ads use "Shop now" and another 20.3% use "Learn more." Four in five ads in your category are making the same ask in the same words.

That last number is the cheapest fix on this page. One in eight food and beverage ads spends money to earn a click and then drops the buyer on a homepage, asking them to re-find the thing the ad just sold them.

Most brands are working from a theory of their buyer that was formed at founding and never tested. It's usually not wrong so much as incomplete: it captures the motivation of the earliest, most enthusiastic customers, who are the least representative people you will ever sell to.

Three sources get you most of the way. Interviews with recent buyers, close enough to purchase that they can still reconstruct the decision. Surveys sized to tell you which motivations are common rather than merely vivid. And conversations with people who considered you and bought something else — the group almost nobody talks to, and the one holding your actual objection.

What comes back is rarely what the brand deck says. In the Cōpow research, the barrier to local organic food turned out to be convenience, not values — which reframed the entire go-to-market. In Meatworks, research surfaced two distinct buyer profiles the existing ads spoke to neither of.

Then the finding becomes the brief, the brief becomes the creative, and the questions above stop being guesses. That sequence — research, then brief, then media — is the whole of what we do.