Where should the next dollar go?

Every paid media dashboard is built to answer the same question: what happened last week?

Spend, impressions, clicks, cost per lead, cost per acquisition. The numbers are real, the dashboards are pretty, and the weekly meetings are full of people nodding at them. And then someone asks the actual question — okay, so what should we do differently next week? — and the room goes quiet.

This is the gap between measurement and decision-making, and most marketing teams live inside it.

I've spent 20+ years on the operator side of this problem. At peak I was running $1M+/month in paid media for client accounts, and the hardest question was never "did last week work?" It was "where does the next dollar go?" And the dashboards were useless for that, because they're built to report, not to decide.

Here's the decision framework I've actually used — not the one I've written in a deck for a client, but the one that runs in the back of my head when I'm staring at a paid media account on Monday morning.

The three questions under the question

When someone asks "where should the next dollar go," they're really asking three questions stacked on top of each other:

1. Where is current spend getting wasted? Which campaigns, placements, or audiences are burning budget with nothing to show for it? This is the easy layer — every account has some of this, and a competent audit surfaces it in a week.

2. Where is current spend working but saturated? Which campaigns are hitting the ceiling of what that channel can produce at acceptable economics? Pouring more money in won't help — CAC will start rising faster than volume. This is harder to see because the campaigns still look "good" on the dashboard.

3. Where is there uncaptured demand we're not touching yet? Which audiences, keywords, or channels are we ignoring that could produce incremental pipeline at reasonable economics? This is the hardest layer because it requires imagining a counterfactual — what would happen if we did try this — rather than analyzing what did happen.

Most dashboards show you question 1. A few show hints of question 2. Almost none help with question 3. And question 3 is where the growth lives.

The saturation problem, specifically

Let me go deeper on question 2, because it's the one most teams get wrong.

Here's the scenario: your Google Search campaigns are running a CPA of $85, well under your $120 target. The CFO looks at that and says "great, let's double the budget." You double it. Three weeks later, CPA is $140. Over target.

What happened? You hit saturation. The highest-intent searches — people literally typing your category into Google — are a finite pool. Doubling spend didn't double the pool; it just pushed you into less-qualified variations of the same query at higher bid prices.

The dashboard told you the first campaign was working at $85. It didn't tell you that working was a ceiling, not a floor.

The way you see this before it happens is by tracking marginal CAC, not blended CAC. Blended CAC (total spend / total conversions) smooths over the saturation cliff. Marginal CAC (cost of the last $10K of spend) tells you whether the next dollar is still buying what the first dollar bought.

Most teams don't track marginal CAC because their attribution doesn't support it. Which brings me to the real point.

The infrastructure problem underneath the question

The reason most teams can't answer "where should the next dollar go" is that their data can't answer it.

If your attribution is last-click-only, you can't see platform substitution — when budget moves from one channel to another, the shift in credit isn't real shift in causality, it's just reallocation of who gets counted.

If your CRM and your ad platforms report different conversion counts, every cross-channel comparison is suspect.

If your lifetime value data lives in one system and your acquisition data lives in another, you can't calculate channel-level ROI, only channel-level CAC.

These are all infrastructure problems, and they all need to be solved before "where should the next dollar go" becomes answerable. Which is why my consulting work starts with infrastructure and not strategy. You can't optimize what you can't see.

What I'm building

This is also the question I built Campaign Budget Optimizer to answer. CBO connects to Google Ads, Meta, Microsoft, LinkedIn, GA, and TikTok, runs the kind of analysis I'd do manually across those platforms, and surfaces where the next dollar should go — not just "what happened last week." Launching May 2026 at campaignbudgetoptimizer.com.

But even without the tool, the framework works. Next Monday, when you're looking at your dashboard, don't ask "did last week work?" Ask the three questions:

1. Where is current spend getting wasted?
2. Where is current spend working but saturated?
3. Where is uncaptured demand we're not touching yet?

Then make the infrastructure investments that let you answer them. The dashboard you inherited isn't going to do it for you.