Free PPC

Toolkit

The account is profitable, the client wants to double the budget, and the temptation is to just lift the cap. Half the time that breaks the account, because the next dollar of spend buys lower-quality traffic by definition. This is the readiness gate and the lever order I run before touching the budget.

The framework

Three gates before any budget change. Pull levers in risk order, one at a time. Read marginal CPA, not average.

Readiness gate (all 5 required)

Demand ceiling state

Lever order (one at a time, never combined)

Marginal CPA, not average

A CPA that goes from $40 to $48 looks like a 20% loss. But if conversions went from 100 to 145, the 45 incremental conversions cost $83 each. Profitable if LTV is over $83. Unprofitable if LTV is under. Average alone hides the truth.

What this framework solves

Scaling is the most common way to destroy a profitable Google Ads account. Every account operates within a finite market of people already searching, and the economics of capturing the next slice are always worse than the slice you already have. The job is not to push past the ceiling, it is to recognise where the ceiling actually is, and to scale up to it without overshooting into unprofitable territory.

The core idea

Google Ads is a demand capture channel, not a demand creation channel. The most efficient customers are captured first, so each additional dollar of spend pursues less efficient traffic by definition. Scaling is non-linear: 2x spend typically yields 1.5x to 1.8x conversions, with conversion rate degradation of 10 to 20 percent from traffic dilution. This means marginal CPA at 2x spend is 25 to 60 percent higher than blended CPA. Average CPA hides this; marginal CPA reveals it. Every scaling decision reframes from “is average cost acceptable” to “is the cost of the next customer acceptable against LTV.”

How to apply it

1. Pass the readiness gate. Five conditions must all be true before any budget increase: CPA or ROAS at or below target for 2+ consecutive weeks, week-over-week variance below 20 percent, 30+ conversions per month, conversion tracking verified, and the business confirmed it can absorb more volume.
2. Decompose impression share. Lost IS to budget on profitable campaigns means scale. Lost IS to rank means quality or bid problem, and budget alone will not fix it. 85 percent IS with no losses means you are at ceiling.
3. Classify the demand ceiling state. Room to grow, scaling gets expensive, near the ceiling, or demand is the constraint. Each state requires a completely different conversation, and only the first one is a Google Ads tactics question.
4. Pull levers in risk order. Budget increase on a constrained profitable campaign first (0 to 5 percent CPA increase). Target relaxation second. Match type expansion third. New campaign types fourth. New channels last. Never combine two levers in one change, or you cannot diagnose the result.
5. Scale in phases, not jumps. 10 to 20 percent increments for conservative scaling, 30 to 50 percent only when CPA is well below target with consistent volume. Stabilize for 14 days at each new level.
6. Calculate marginal CPA, not average CPA. Average going from $40 to $48 looks like a 20 percent loss. But if conversions went 100 to 145, the 45 incremental conversions cost $83 each. If LTV is $200, that is profitable growth. If LTV is $60, $48 average is the ceiling.
7. Define kill criteria in writing before you increase. “If CPA exceeds $60 for 10 consecutive days, revert to previous budget.” Pre-defined, time-bound, communicated to stakeholders, enforced without negotiation.
8. Check operational capacity for lead gen. Effective CPA equals reported CPA divided by close rate. If close rate drops from 20 to 12 percent because the sales team cannot keep up, effective CPA increases 67 percent even if Google Ads CPA looks unchanged.

When to use it

Use this whenever a client asks “can we double the budget” or before executing any spend increase above 20 percent, especially in lead gen and high-CPC verticals where the cost of scaling the wrong way is immediate and large. It also applies in reverse: when an account looks like it has stopped growing, this framework distinguishes a management problem (“we have not optimized enough”) from a structural one (“we have captured most of the available demand”). It does not apply when the account fails the readiness gate. In that case, fixing efficiency comes first, and scaling is a later conversation.