Food and supplement brands that sell direct to customers have the best margins in e-commerce. They also have one of the worst data problems. Your pixel misses about a third of purchases.
30 min. We read your numbers, not a deck. No obligation.
Returns come in after the ad platform stops counting. Targets get set against gross revenue, not what you keep. So Meta builds a picture of your customer that is too generous. The profit you thought you were making never lands in your account. It exists only in the platform’s report.
Returns are invisible to the algorithm. A customer buys a probiotic bundle. The pixel fires on purchase. Two weeks later they return it. The ad platform never gets that signal. So it keeps chasing customers who look exactly like that returner.
Subscription renewals vanish. Your real profit comes from the customer who reorders for six months. Not the one who tries once and leaves. But renewal events from your payment system are almost never tracked cleanly. So the algorithm learns from first orders only. It bids for trial-seekers instead of loyal buyers.
Meta’s February 2025 health and wellness restriction. Some supplement brands lost the ability to optimize for purchases overnight. If your account was hit, the algorithm now learns from a fraction of what it used to see. That comes on top of the data loss that was already there.
Your goods costs move. Ingredient prices, supplier lead times, and packaging reorders mean the margin this quarter may not match last quarter. That moves your contribution profit — what’s left after you subtract returns, VAT, product costs, and ad spend from your revenue. It’s the number your bank account actually keeps. Recalculate it every month, or you are steering by a snapshot from a quarter when the business looked different.
Mindlink’s founder built Cebulka Catering — a subscription diet-delivery brand in Warsaw — to PLN 502,325 in revenue across nine weeks, then sold it. A food operation has daily production costs, weekly renewals, and churn that shows up before the month closes. Run one of those, and you learn fast what real profit means when there is no gross revenue line to hide behind.
That is not agency experience borrowed and pointed at food brands. It is operator experience from inside the unit economics.
The team has also shipped tracking and performance campaigns for Volkswagen, Audi, KFC, and WizzAir. On Chocolissimo — a multi-market personalized chocolate store — fixing the measurement alone cut the cost to acquire a customer by 22%. Same creative. Same budget. Same targeting. The only thing that changed was how much of the real sales the algorithm could finally see.
Step 1 — The diagnosis is free. You share your screen. We check three things together. How many confirmed orders Meta never recorded. Whether your targets are set against gross revenue or what you actually keep — the gap is often €8,000–15,000 a month in decisions pointed the wrong way. And what a customer really costs to acquire once returns come back. If there is no real gap, we say so, and you lose 30 minutes.
Step 2 — The Profit Foundation Audit (€5,000, three weeks). We match up your ad data, website analytics, and order system across 12 months. We work out the profit you keep per order, using your real product costs, shipping, and return rate. We map which ad angles bring in customers who reorder and which bring in one-time buyers.
Step 3 — The Profit Proof System (ongoing). We take over Meta and produce at least 50 fresh ads a week across 10–15 angle families — mechanism, transformation, social proof, problem-solution, ingredient story, repeat-buyer loyalty, and more. Each ad ships in both square and vertical formats.
In supplements and functional food, the ad that wins a first-time buyer is not the ad that wins someone already on a competitor’s product. And neither one wins back a customer who lapsed. Ten to fifteen genuinely different angle families give the algorithm enough variety to find all three. Most agencies ship three to five ads a month — one angle, a few sizes. The algorithm runs through that in days, then has nothing new to learn from.
Fresh ads are what scaling needs. So we ship at least 50 a week — and the production fee simply covers what they cost to make. Gemini will render you an image for cents. So will we. You are not paying for pixels. You are paying to know which fifty to make.
The second part is how we earn. We take a share of the new profit we create. Profit here means the cash left after VAT, returns, product costs and ad spend — we call it contribution profit. If that number does not grow, we earn nothing on that side.
30 minutes. We read your numbers, not a deck. No obligation.
It depends on two things. How Meta classified your account, and whether you asked Meta to move it out of the restricted category. We check this in the free call. Maybe the restriction cannot be lifted. Maybe Meta will not let the algorithm optimize for purchases. Then our model may not fit. We will say that plainly, rather than take a client we cannot help.
We recalculate the profit you keep every month. We use your real costs, not a fixed percent we guessed at the start. If your ingredient costs moved, the real cost of acquiring a customer moved too. We track it.
Yes. We read the confirmed order data straight from your order system or payment processor and send it to the ad platforms from the server, not the browser. The pixel does not need to fire. The confirmed transaction is the source of truth.
Yes. Taste, texture, and sensory proof that cannot be delivered through a screen mean social proof and transformation angles carry more weight than product feature detail. We build creative strategies around what drives repeat purchase in the specific category — supplements, functional food, catering subscriptions — not a generic e-commerce template.
We put your ad account next to your real order data and name the gap. How many purchases the platform missed. What profit you actually keep after returns, VAT, goods, and ad spend. And where the money you can still recov
30 minutes. We read your numbers, not a deck. No obligation.