Marketing budgets have been flat since 2022 and now sit at 7.8 percent of company revenue, 18 percent below where they were four years ago, according to Gartner's 2026 CMO Spend Survey. Fifty-six percent of CMOs told Gartner they do not have the budget to deliver this year's strategy.
When the pie stops growing, one question decides everything. Where does the next dollar go? And the number most teams reach for to answer it was never built to.
The Stealth Tax on Your Budget
8.51% | $63B | ~ 40% |
|---|---|---|
Invalid Paid Traffic Roughly one in twelve clicks is not a real buyer. | Lost to invalid traffic In global ad spend went to invalid traffic last year. | Rise in search CPCs Over three years. The flat part of the curve keeps getting pricier. (industry, 2026) |
Source: Media Post
Before you optimise a single channel, this is what the next dollar is walking into.
There Are Two ROIs. Your Dashboard Only Shows the One That Flatters You
Average ROI is everything a channel ever returned divided by everything you ever spent on it. It is the number in the QBR deck, the one that has survived three budget reviews, the one your team defends like a family name.
Marginal ROI asks a colder question. What will the next dollar return?A
On a customer call earlier this year I said something that took everyone by shock. They kept pouring money in Linkedin which was already saturated. The ROI you earned up to today is not the ROI your next dollar will get.
I have said a version of that sentence to more teams than I can count. The channel changes, but the pattern never does.
Every channel runs on a response curve. The first dollars buy pipeline cheaply, because they reach the people who were easiest to move. Each dollar after that reaches someone slightly harder to convince. Eventually the curve flattens, and everything you spend past that point buys mostly comfort. The tenth thousand dollars into a channel almost never works as hard as the first.
It is a structural trap. Your average ROI was harvested on the steep part of that curve. Your next dollar lands on the flat part. The dashboard proudly reports the harvest, the budget decision happens at the flat end, and nothing on the screen tells you those are two completely different places.
The two-ROI curve:
Only 32 percent of marketers worldwide measure spend across both their digital and traditional channels, according to Nielsen. Two thirds are reading half a map and calling it a strategy.
Saturation, Lag, and One Dead Deal: How the Average Lies to You
The average flatters your best channel in three specific ways:
The first is saturation, the curve I just described. Past a point, more money into the same channel simply buys worse and worse returns, even while the lifetime average still looks glorious.
Measure the lag. Paid channels carry momentum from prior quarters. In one model review this year, a single channel was projected to throw off roughly $9 Million in pipeline in a quarter with zero new spend, running entirely on money spent months earlier. That is the trap : channel can look alive on the dashboard while your current dollars into it do almost nothing, because the pipeline arriving today was bought a long time ago. It runs in reverse too. Go dark now and the bill lands next quarter.
Spot the outliers. Mix models are statistical machines, and one enormous deal can bend an entire channel's curve. We recently watched a single seven-figure opportunity, already closed and lost, make a minor channel look like a breakout. One dead deal, quietly steering a live budget. It got caught only because someone asked a simple question. What happens to this recommendation if we pull that one deal out of the math? Ask that about your biggest channel this week and see if the story survives.
Now widen the lens, because everyone is making the same move at the same time. We are all piling onto the same curves, which means the flat zone arrives sooner for everyone in the market, not just you.
Never Take an Increase List Without a Decrease List
So price the next dollar before you spend it. Three questions, every time, before any budget moves.
What did the last dollar return, not the lifetime average? If nobody in the room can answer, your allocation is running on reputation, not performance.
How much of this quarter's number is riding on last quarter's spend? Separate the momentum from the actual performance before you reward either one.
Is a single deal bending this curve? Pull the largest opportunity out of any channel analysis and check whether the case still stands.
Then adopt the one rule I now push on every team I sit with. Never accept an increase list without a decrease list. If your analyst, your agency or your shiny new AI can only tell you where to add money, that is not optimisation. That is cheerleading. Paired recommendations are the only ones you can actually check: move real dollars in both directions across a handful of campaigns, then score the projection against what happened. Small tests, both directions, every quarter.
ADD ↑ | CUT ↓ |
LinkedIn Sponsored ↑ Bing Search ↑ Podcast Sponsorship ↑ | Google Display ↓ Facebook Retargeting ↓ Programmatic Ads ↓ |
One recommendation, or it does not ship.
Advice you cannot falsify is not intelligence. It is theatre with a confidence interval.
This newsletter is called Moneyball CMO for a reason. In 2002, the Oakland A's baseball team had a problem every CMO recognises: a fraction of the budget their rivals had, and the same results expected of them.
The New York Yankees were spending $126 million on players. The A's had $40 million. Instead of copying what the big spenders did, the A's asked a different question: what does a win actually cost at the margin, not on average? They stopped paying for famous names and started pricing each player by what they would contribute to the next win specifically.
The result: 103 wins, identical to the Yankees, at a third of the cost. The lesson for marketing is direct. Your best channel is your Yankees payroll. It looks impressive. It has a great track record. And it is almost certainly overpriced for what it will return from here.
Choose the Price Over the Reputation
The best performing channel in your board deck is not lying to you on purpose. It is just showing you the flattering number, the one built on the steep part of a curve your next dollar will never touch again. Nothing on the screen tells you those are two different places.
Average ROI is a reputation. Marginal ROI is a price.
P.S. Most teams cannot see where their channels saturate, because no dashboard draws the curve. Drawing that curve, channel by channel and campaign by campaign, is exactly what RevSure's marketing mix models do. If you are second-guessing where your next dollar belongs, reply and I will show you what a response curve reveals about a budget shaped like yours.