Judge The Portfolio, Not The Campaign
How to run an exec-grade marketing org when outcomes are uneven
Most marketing teams get managed like they’re running a factory: every campaign is expected to produce a predictable output, every month should “perform,” and every channel must justify itself in isolation.
But marketing behaves more like a lottery, a portfolio, a power law: most bets return modest results, a few miss entirely, and occasionally one hits so hard it changes the quarter.
Rory Sutherland has a simple idea that most orgs hate: marketing works more like Hollywood than manufacturing. Watch him explain why marketing budgets should behave more like film studios than factories:
Most releases do fine. Some flop. A few become breakout hits.
The value of the category comes from the rare outsized wins, not the average outcome.
So when finance, procurement, or agencies force marketing into a Ford-style production line, they punish the variance that creates the upside.
In other words: they demand certainty and then complain marketing has no magic.
This is where most Marketing leaders get stuck.
You know the truth, but you still have to run the org inside a company that rewards predictability.
So here’s Rory’s insight as an operating model you can actually use.
The real job: manage variance, don’t eliminate it
“Marketing is a cost center” is often code for: “We don’t understand how you create value, so we’ll manage you like a machine.”
Your job as a marketing leader is to make the system legible without killing its upside, especially when your own pipeline timeline is tighter than your exec team admits.
That means three things:
Separate predictable growth from optional upside
Create guardrails that finance can trust
Measure the portfolio, not each campaign
If you do this well, you get two wins:
You keep credibility with execs
You keep room for breakout results
The portfolio model: Core, Growth, Breakout
You need a budget story that makes sense to CFO brains.
Use three buckets:
1. Core (low variance)
The stuff that reliably delivers.
Proven channels, proven offers, proven audiences
The goal: efficient volume, steady pipeline, steady revenue assist
How it gets measured: efficiency and consistency.
2. Growth (medium variance)
Scaling what works and expanding reach.
Doubling down on winners
New audiences adjacent to what’s already converting
Distribution improvements, lifecycle, conversion work
How it gets measured: scalable lift and repeatability.
3. Breakout (high variance)
This is where the “lottery tickets” live.
Bold creative bets
New channels
New category narratives
Weird experiments that might look dumb until they work
How it gets measured: speed to signal, learning velocity, and upside potential.
Here’s the key line you need to adopt as a leader:
Core pays the bills. Breakout buys options. Growth turns options into systems.
Most teams fail because they blend all three, and then judge everything by “did it hit this month.”
Guardrails: how to keep variance without looking reckless
Variance doesn’t mean chaos. It means rules.
Here are four guardrails that make execs calmer while keeping your upside intact:
Guardrail A: Protect the Breakout budget
Pick a percentage and defend it.
Early stage teams: 10–15%
Mature teams: 5–10%
And make one rule: Breakout cannot be raided to cover Core misses.
If you allow that, you train the org to kill upside when results get tense.
Guardrail B: Kill fast, scale faster
Breakout should be brutal.
Fast tests
Clear signals
Hard stop dates
And when you get a hit, you do the opposite:
Fund it properly
Move it into Growth
Build repeatability
You’ve seen this pattern: a dozen tests create small lifts or nothing at all, and then one weird idea becomes the pipeline story your CRO repeats in QBRs.
Guardrail C: Measure the portfolio
A single campaign failing doesn’t mean marketing failed.
If your Breakout bucket is designed to produce 1–2 meaningful winners per year, you should expect many misses. That’s the cost of upside.
The goal is not “every bet wins.”
The goal is: the portfolio wins.
Guardrail D: Report leading indicators, not hopes
Finance hates fluff. So stop reporting fluff.
For Breakout, report:
time-to-signal
cost per signal
tests launched per month
share of attention (where relevant)
qualified response lift (not just clicks)
If you can’t define a signal, you don’t have an experiment. You have content.
The “credit problem” and why it kills breakout work
Rory also points out something leaders feel in their bones:
When a breakout win happens, marketing rarely gets full credit.
But when variability shows up as misses, marketing gets punished immediately.
So teams rationally adapt:
fewer risky bets
more safe, incremental campaigns
a culture of “don’t get fired”
That’s how brands slowly become interchangeable.
Your job is to change the incentive structure:
Protect smart risk
Punish slow learning
Reward speed and signal
When a Breakout campaign drives a surge in demo requests, sales gets credit for “better outreach.”
When a test underperforms, marketing gets asked “what went wrong.”
This asymmetry trains teams to avoid anything that can’t be explained in a single slide, the same dynamic that makes hiring and org design drift toward “safe resumes” over real operators.
It’s why I like using a clear scorecard like the one in Fractional CMO vs Marketing Director: The Scorecard.
What to say to finance in one sentence
If you only steal one line from this post, steal this:
“Judge the portfolio, not the campaign, and I’ll show you the guardrails that prevent waste.”
That framing shifts the conversation from “prove every campaign” to “prove the system.”
And that’s where marketing leadership actually lives.
The resource that makes this real
In the paid section below, you’ll get a one-page downloadable tool you can use today:
The Variance Budget + Portfolio Scorecard (Excel Sheet)
It forces your spend into Core/Growth/Breakout, builds kill rules, and gives you a CFO-friendly monthly report view.
It’s the fastest way to stop the “factory output” conversation without sounding like you’re making excuses.
Get the tool that lets you defend variance without sounding like you’re making excuses, and turn your next budget conversation from interrogation to strategy session.
Paid subscribers also get weekly curated remote US marketing jobs ($50K–$250K) posted over the last 7 days — 73 roles this week.
If you want to run this model without building it from scratch, the Excel Scorecard below is the exact template.