How to reallocate a fixed budget under a "Grow CAC" mandate

The short answer

When leadership says "grow paid spend 25% next quarter," the naive move is to scale every channel 25%. That reliably buys inefficient volume, because it pours money into channels that are already saturated alongside the ones that still have room. The disciplined move is a revert–grow–reserve structure: revert the channels that were buying volume at a loss to their prior spend, grow the genuinely efficient channels, and hold the remainder as a governed reserve instead of force-deploying it. This protects your CAC target while still funding growth — and it turns a fixed budget into a decision framework rather than a static plan. I built exactly this for a delivery marketplace facing a large quarterly budget and a mandate to grow spend without blowing out CAC.

Why "scale everything 25%" fails

A flat scale assumes every channel has the same marginal return. It never does. Some channels are still on the steep part of their response curve — an extra dollar there earns. Others have saturated — an extra dollar mostly inflates cost with little incremental volume.

On the account I modelled, the "do nothing but scale" path produced a blended CPA about 40% above target. The budget mandate itself was outpacing the market's demand growth, so pushing harder on saturated channels was buying progressively more expensive customers. Flat scaling would have hit the spend target and missed the efficiency target — which is the outcome that gets growth leads fired.

Step 1: Tie out the data before you model anything

Before allocating a dollar, reconcile the numbers. On this account I caught a ~$0.6M discrepancy in the source data before modelling. Nothing gets allocated on numbers that do not reconcile — because every downstream decision inherits the error. This is unglamorous and non-negotiable. A reallocation model built on unreconciled data is confident and wrong.

Step 2: Find where marginal spend still returns

Fit a response curve per channel — a log-log marginal-return model works well — to find where each channel sits on its curve. The question is not "which channel has the best average ROI" but "which channel still returns on the next dollar." Those are different, and the difference is the whole game. Average ROI tells you where you have been efficient; marginal return tells you where the next dollar should go.

This splits your channels cleanly into: still-returning (grow these), and saturated/loss-buying (do not scale these — revert them).

Step 3: Revert, grow, reserve

The structure:

• Revert the saturated, loss-buying channels to their prior-quarter spend. On my account these were the channels scaling volume at a loss — reverting them stopped the bleed immediately.
• Grow every efficient channel by the mandated increase. This is where the growth money actually goes to work.
• Reserve the difference. Reverting the losers freed a meaningful pool. Instead of forcing it back into channels that could not absorb it profitably, I held it as an uncommitted reserve — on this account, close to $1.8M.

The reserve is the key insight. A fixed budget does not have to be fully committed on day one. Holding part of it, governed, lets the plan respond to in-quarter signal instead of locking spend in months before you know what is working.

Step 4: Govern the reserve with a rule, not a vibe

An uncommitted reserve only works if it is governed, or it leaks into business-as-usual and quietly disappears. The operating rule I use has four parts, stated in this order:

1. Gate — the specific condition that must be met to release reserve funds.
2. Cap — the maximum that can be released per release event.
3. Endpoint — a defined point at which unreleased reserve stops being available, so it cannot drift into general spend.
4. Fallback — what happens to reserve that is never released.

Gate, cap, endpoint, fallback. State the rule first, then attach the numbers. A reserve without this structure is just underspend with a nicer name.

Step 5: Prove reallocation decisions with incrementality, in two layers

Reallocation decisions cannot rest on last-click ROI — that just moves budget toward whichever channel is best at claiming credit. I built a two-layer incrementality framework:

• Layer 1 (within channel): structured A/B lift tests across four universal levers — Audience, Creative, Bidding, and Landing Page — run per channel. This tells you how to improve each channel.
• Layer 2 (across channels): geo or holdout tests that validate whether channel-level reallocation actually caused incremental volume. This tells you whether to move money between channels.

Layer 1 optimises within a channel; Layer 2 justifies moving money between them. Together they mean every reallocation rests on measured lift, not attribution artefacts.

The model in one line

Tie out the data, fit marginal returns, revert the losers, grow the winners, reserve the rest under a gated rule, and validate with two-layer incrementality. That converts a blunt "spend more" mandate into a system that protects the target while funding growth.

The bottom line

A "grow spend 25%" mandate is not a growth plan — it is a constraint. The teams that hit both the spend target and the efficiency target are the ones that treat a fixed budget as a decision framework: grow where the marginal dollar returns, revert where it does not, and hold a governed reserve for what the quarter reveals. The budget is the input. The allocation rule is the work.

If you are facing a budget mandate that threatens your CAC target, that reallocation modelling is core to how I work with growth teams — see DTC paid media consulting. You can also pressure-test your current allocation with the DTC Paid Media Audit Checklist.

---

Frequently asked questions

How do I grow ad spend without increasing CAC? Do not scale every channel equally. Grow only the channels where the marginal (next-dollar) return still clears your CAC target, revert saturated channels to prior spend, and hold the remainder as a governed reserve. Flat scaling inflates CAC because it feeds saturated channels alongside efficient ones.

What is a budget reserve and why hold one? A reserve is the portion of a fixed budget you deliberately leave uncommitted, released only when a defined condition is met. It lets you respond to in-quarter performance instead of locking all spend in advance — but it must be governed with a gate, cap, endpoint, and fallback or it leaks into general spend.

Should reallocation decisions use last-click ROI? No. Last-click moves budget toward channels that claim credit, not channels that cause incremental volume. Validate cross-channel reallocation with geo or holdout incrementality tests, and use within-channel A/B lift tests to optimise each channel individually.