Key takeaways
- There is no universal dollar figure — your budget is a function of gross margin, MER, and a break-even ROAS, not a number you copy from a competitor.
- As a rough orientation, DTC Meta budgets often land somewhere in 10–30% of revenue — a starting range, not your number. Break-even ROAS = 1 ÷ gross margin.
- Reserve 10–20% of spend for testing. It’s how you find the next winning creative before the current one dies.
- Scale by holding blended MER steady as you increase spend, not by chasing each campaign’s in-platform ROAS.
“How much should we spend on Meta?” is the question every DTC founder eventually asks, and the honest answer is that the dollar amount is the wrong thing to fixate on. A $5K/mo brand and a $500K/mo brand can both be spending exactly the right amount or wildly the wrong amount. The figure only means something relative to your margins and your math.
So instead of a number, here’s the framework Kova uses to set and grow a Meta budget, and the same one it brings to a paid media engagement.
Step 1 — Start with break-even ROAS, not a budget
Before you decide how much to spend, you need to know the point at which a sale stops making money. That’s your break-even ROAS, and it comes straight from gross margin:
If your gross margin is 40%, you break even at a 2.5× ROAS on that order. A 70%-margin brand breaks even at ~1.4×. This single number tells you whether a campaign is actually profitable, and it’s why two brands with the same ROAS can have completely different fates.
Step 2 — Budget as a percentage of revenue
Once you know break-even, budget becomes a lever, not a fixed cost. As a rough orientation, most scaling DTC brands land somewhere in 10–30% of revenue on Meta, depending on stage and ambition. Two broad postures:
- Early / aggressive growth: often 20–30%+ of revenue, deliberately running closer to break-even to buy market share and first orders.
- Established / profit-focused: often 10–20%, optimizing for contribution margin over raw top-line.
The percentage isn’t the goal. It’s the output of a decision about how hard you want to grow versus how much profit you want to bank this quarter.
Step 3 — Carve out a testing budget
Whatever your total, reserve 10–20% for structured testing: new creative, new hooks, new audiences. New brands skip this and wonder why performance flatlines. They’re spending 100% on proven ads that are quietly fatiguing, with nothing in the pipeline to replace them. Testing isn’t waste. It’s how you manufacture your next winner on schedule instead of by luck, which is the same logic behind shipping new concepts every month and testing them against the last.
Step 4 — Judge it on blended MER, not platform ROAS
As you scale, the number that matters is MER (marketing efficiency ratio): your total revenue divided by total ad spend across everything. Platform-reported ROAS over-counts, especially on retargeting, so it’s a directional input and never the verdict. A blended MER of 3–4× is healthy for many brands, but a high-margin brand with strong repeat purchase can profitably run lower, because lifecycle revenue pays back the thin first order.
This is also why clean measurement matters before you scale a dollar. If your tracking is lying, every budget decision downstream is built on sand.
Step 5 — Scale in steps, hold the ratio
When the math works, scaling is simple in principle: increase spend in measured increments (a common step is ~20–30% at a time), and watch whether blended MER holds. If it holds, push again. If it slips, you’ve found the current ceiling, and the fix is usually fresh creative or a new offer rather than more budget jammed into a fatiguing ad. Why that happens is covered in why your ROAS drops when you scale.
So how much should you spend?
Enough to test relentlessly, capped by the point where blended MER falls below the threshold your margins can support. For most DTC brands that’s a number between 10% and 30% of revenue, but now you can derive your number instead of guessing it. Spend is a dial you set with math, and then turn up as the system proves it can hold.