Cross-Brand Budget Allocation for Paid Media

The brand with the highest ROAS is not automatically the one that should get more money. The brand whose owner emailed you twice this week is definitely not. This piece is about building a decision process that survives contact with five brands, three channels, and a budget that never quite covers everyone’s ambitions.

Why the obvious answer is wrong

The instinct is to fund your best performer. Brand A runs at 4.2x, Brand C runs at 2.1x, so more money goes to Brand A. Clean, defensible, and usually wrong.

It’s wrong because average ROAS tells you where a brand has been, not where the next dollar goes. Brand A might be running at 4.2x precisely because it’s under-spent — harvesting its cheapest, highest-intent demand and nothing more. Push another $20k into it and you’re buying progressively colder audiences, and that 4.2x collapses toward 2.5x on the incremental spend even though the blended number still looks healthy. Meanwhile Brand C at 2.1x might have a wide-open mid-funnel that returns 3x on the next dollar because nobody’s funded it yet.

What you actually care about is marginal return: the return on the next increment of spend, per brand, not the average return on all spend to date.

Allocation decisions made on averages systematically overfund saturated winners and starve brands that still have cheap demand left to buy. If you take one idea from this piece, take that one.

The second reason the obvious answer is wrong: platform ROAS isn’t additive across a portfolio. Meta reports its number, Google reports its number, each measured by the system asking you for more budget, each double-counting conversions it wants credit for. You cannot stack six platform ROAS figures into a portfolio view. A real portfolio number comes from your commerce data — Shopify or WooCommerce orders — divided by total spend across every channel and brand.

The three questions every allocation decision answers

Before any framework, every dollar-level decision is really answering three questions in order. Get the order wrong and the framework won’t save you.

• Is this brand still learning or is it saturated? A brand that hasn’t hit its efficient frequency ceiling has room to grow profitably. A brand already saturating its addressable audience will convert new dollars at a worse rate no matter how good its creative is.
• What does the marginal dollar return, and how fast does it pay back? Marginal return and CAC payback are two different guardrails. A brand can show strong marginal ROAS while its payback stretches to fourteen months — fine if you’re capitalized for it, a cash-flow trap if you’re not. Both numbers gate the decision.
• What does the portfolio need strategically? Sometimes the highest-return dollar isn’t the right dollar. A brand you’re preparing to sell, a launch that needs a demand floor, a seasonal brand heading into peak — these override pure efficiency for a defined, written-down reason. The key word is written-down.

A decision framework you can actually run

Portfolio-level rules like the 70/20/10 split — 70% to proven bets, 20% to scaling contenders, 10% to experiments — are a useful mental model for how much risk your total pool should carry. But they don’t tell you which brand is a proven bet this month. For that you need to score brands against each other on the same axes, every cycle. Score each brand on these signals, then let the score drive the move:

• Marginal ROAS — return on the next increment, not the average. Reallocate toward brands above your portfolio blended target; away from brands below target and falling.
• Saturation — frequency and audience penetration versus addressable. Reallocate toward brands with room left before the efficient frequency ceiling; away from brands already saturating the audience.
• CAC payback — months to recover acquisition cost. Reallocate toward brands inside your cash-flow tolerance; away from brands stretching past it.
• Pacing — actual spend versus planned shape. Reallocate toward brands under-pacing with healthy efficiency; away from brands over-pacing into weak return.
• Strategic weight — launch, seasonal peak, exit prep. Reallocate toward a documented strategic priority; away from brands with no strategic reason to protect them.

The scoring doesn’t have to be sophisticated. A simple weighted rank — marginal ROAS and CAC payback carrying the most weight, strategic weight as a manual override — puts your brands in an order every cycle. The brand at the top gets the incremental dollar. The brand at the bottom funds it.

The rules that keep it from breaking

A scoring model without guardrails will happily strip a brand to zero and blow up another brand’s learning phase in a single move. Wrap the model in these five rules:

• Set a floor and a ceiling per brand. No brand goes to zero — you lose the learning signal and the platform algorithms reset. No brand takes more than a defined share of the pool — concentration is its own risk, and platforms get less efficient at the top of a spend curve anyway.
• Move in bounded increments. Cap any single reallocation at roughly 20–25% of a brand’s current budget. Bigger swings throw Meta and Google campaigns back into the learning phase, and the efficiency you were chasing evaporates while the algorithm re-stabilizes.
• Reallocate on a fixed cadence. Weekly or biweekly, on the calendar, not when someone complains. The cadence is what converts allocation from a political negotiation into an operating rhythm.
• Gate every move on two numbers, not one. Marginal return says a move is attractive; CAC payback says you can afford it. A move needs to clear both.
• Write the reason down. One line per move. It’s the only way you find out three months later whether your reasoning was any good, and it’s what lets you defend the exit-prep brand you’re deliberately funding below its efficient return.

At three brands you can run this in your head and a spreadsheet. At five or six, running it every week — pulling marginal curves per brand, reconciling blended spend against real orders, logging every move — is a standing part-time job. That’s the point where operators either hire for it or automate it. The prerequisite is clean structure underneath: managing ad accounts across multiple brands is what makes a portfolio view possible in the first place.

Worked example: five brands, one pool

Say you’re moving $15k of incremental monthly budget across a five-brand portfolio. The naive move funds Brand A because it posts the highest average ROAS. The framework move looks different:

Same $15k, completely different destination than “fund the winner.” Nobody in this decision emailed you. That’s the tell that the process is working.

What to stop doing

A short checklist of allocation anti-patterns worth killing on sight:

• Allocating on average ROAS. Covered above — it’s the single most common and most expensive mistake.
• Reallocating reactively. If your budget moves are triggered by complaints, the loudest brand owner is running your media plan.
• Ignoring CAC payback. High marginal ROAS on fourteen-month payback can starve you of cash while the dashboard looks great.
• Moving too fast. Blowing up the learning phase with a 60% swing costs you more efficiency than the reallocation gains you.
• Trusting platform-reported revenue for cross-brand math. It double-counts and it isn’t additive. Blended, order-based numbers only.

Where this leaves you

Cross-brand budget allocation stops being political the moment you can see marginal return per brand and you’ve agreed the rules in advance. The math matters, but so does the discipline — which is why running it on rules and an optimization loop beats running it on whoever complained last.

Cesara was built for this decision. It connects Meta, Google Ads, and TikTok with per-client OAuth, pulls real orders from Shopify and WooCommerce so the return math is blended and honest, and reallocates budget across accounts on a scored, documented cadence instead of on gut feel. See how the product works, or how Cesara fits multi-brand ecommerce operators.