Original source: RevOps FM
This video from RevOps FM covered a lot of ground. 6 segments stood out as worth your time. Everything below links directly to the timestamp in the original video.
What happens when a senior marketer with no math background is handed the problem of proving marketing's value — and actually solves it? The answer changed how he thought about the entire profession.
HP Crisis Pushed CMO to Dig Out University Textbooks — and Build a New Analytics Company
A confrontation with then-HP CEO Mark Hurd over marketing's inability to prove its value sent Mark Stouse home one Friday to excavate old university math textbooks from his garage. A non-mathematician by training, he was searching for a defensible answer to a straightforward question: what is marketing actually delivering? What he found, buried in those pages, was multivariable regression — the same statistical method underpinning climate science and epidemiology — and it became the intellectual foundation for everything that followed. Hurd, to his credit, arranged a mentoring session between Stouse and HP CFO Bob Layman, an experience Stouse later described as transformative enough to thank Hurd for before his death.
The structural problem Stouse diagnosed at HP has never really gone away. Marketing teams at enterprises like Honeywell Aerospace were forced to hire large analytics staffs just to reduce the time lag between data collection and usable insight — an approach that cost tens of millions and remained inaccessible to most companies. That gap between what the math could theoretically reveal and what organisations could practically afford to learn is what ultimately drove him to found Proof Analytics, automating the regression process so that the insights arrive before the decision has already been made without them.
"I either have to do something to fix this or I just need to go do something else — because it's not just about budget issues, it's about credibility. What am I actually doing here?"
Multi-Touch Attribution Is Selling Marketers a Measurement Tool That Cannot Do What It Promises
The central flaw in multi-touch attribution is not technical sloppiness — it is a category error. MTA records effects and calls them causes. It observes that a thousand buyers downloaded an ebook, attended a webinar, and clicked an ad before signing a contract, then distributes credit across those touchpoints according to weighting models that marketing teams themselves configure in advance. The mathematics required to convert that pattern into causal proof simply does not exist within the MTA framework. Stouse illustrates the problem with a sharp example: in a self-serve SaaS product, every new user received a welcome email, so the welcome email appeared endlessly in attribution data as a high-influence touchpoint — because presence and causality are not the same thing. In B2B, where time lags between marketing investment and revenue outcome can stretch two to three years for demand-generation activity and double that for brand, the assumption embedded in MTA that cause and effect are near-simultaneous is not a minor approximation. It is structurally wrong.
The real question is why adoption happened anyway. Stouse's answer is unsentimental: most marketers lacked the mathematical training to recognise the flaw, and MTA was framed as a way to extract insight from data the marketing stack was already generating. It was easy, and easy things get adopted. What most people miss is the downstream consequence — that finance and the C-suite, who often do have quantitative instincts, read MTA outputs and find them unconvincing precisely because the logic doesn't hold. The credibility problem MTA was supposed to solve, it quietly deepens.
"I don't think that marketers understand how transparently untrue the premise of MTA is to everyone else but them."
Market Mix Modeling Works Like a GPS — and Uses the Same Math as Climate Science
Market mix modeling is, at its core, a probabilistic navigation system for business decisions. It takes everything a company controls — ad spend, pricing, promotions — alongside everything it does not control — competitive moves, macroeconomic conditions, seasonality — and calculates their relative causal contributions to a given outcome across time. The output is not a single definitive answer but a ranked view of what has been driving results historically, combined with a forward forecast that updates automatically as new data arrives. The GPS analogy is precise rather than decorative: regression mathematics is literally embedded in the navigation software on a smartphone, recalculating the optimal route in real time as conditions change. Applied to marketing, the same recalculation logic tells a team not just where performance currently stands but whether the gap between forecast and reality is widening — and how much time and resource remains before the destination becomes unreachable.
The structural reality is that most business questions are, at their foundation, navigation problems: where are we, where do we need to go, and do we have enough to get there? What MMM provides that conventional reporting does not is the causal architecture beneath the numbers — the ability to distinguish which investments are actually moving the vehicle forward from those that are merely correlated with the journey.
"Most business questions — and indeed a lot of life questions — are navigation questions: where am I, where do I want to go, what's the best way to get there?"
Brand Spend Doesn't Drive Demand — It Drives Deal Size and Speed, and CFOs Are Missing That Distinction
When a company spends $100,000 a month on display advertising and the CFO demands to know what it is producing, the answer almost never arrives in the currency the question assumes. Marketing teams have historically tried to justify display budgets through the lens of demand generation — clicks, leads, pipeline — and that framing consistently fails because display advertising is not, in most cases, a demand-generation investment. It is a brand-reputation investment, and the outcomes it drives are different in kind, not just in timing. Stouse points to two metrics that brand investment consistently moves when measured correctly: average deal size and average deal velocity. Brand reputation, as he frames it, is grease on the wheel of the deal — it makes existing buyers purchase more than they otherwise would, and purchase faster. In B2B, where buying decisions are high-cost, high-risk, and increasingly burdened by external uncertainty, that acceleration and expansion effect is commercially significant.
The time-lag gap compounds the problem. Brand effects are two to three times more delayed than demand effects, and they persist far longer — their half-life extending well beyond the quarterly cycles finance uses to evaluate spend. Demand investment, by contrast, decays within months. If the conversation with the CFO never explicitly addresses time lag, Stouse argues, the ROI of brand investment is not just unmeasured — it is mathematically unmeasurable. A company that tracks year-two deal size uplift among existing customers, he suggests, is looking at one of the most honest signals of how trusted it actually is.
"Brand reputation is grease on the wheel of the deal — it makes people buy more than they otherwise would, and buy faster than they otherwise would."
A Company Detected COVID's Market Impact Before COVID Was a Story — Through External Data in an MMM Model
Before COVID-19 was publicly recognised as an economic disruption, a company's market mix model began signalling that its historically reliable channel investments — many of them physical events — were forecast to lose effectiveness within six months. The signal came not from internal marketing data but from external macro indicators woven into the model: economic, behavioural, and environmental data available through government sources, universities, and financial institutions. Acting on the forecast rather than on what current performance still appeared to confirm, the team began pulling investment from those channels early, recovering capital before the disruption arrived. When reality matched the model's prediction, the team had a second asset: when finance moved to impose cuts of 30 to 40 percent, the team modeled the three-year downstream revenue impact of those reductions and presented the findings. The projected damage was severe enough that finance accepted a substantially smaller cut and found the remaining savings elsewhere.
What this story illustrates is not simply predictive accuracy — it is the structural advantage of treating analytics as a governance instrument rather than a reporting function. The model did not just describe what had happened; it created negotiating leverage against a decision that would have been genuinely harmful. The external data required to produce this kind of signal is, Stouse notes, largely free — sourced from government databases, central banks, and academic institutions — which removes one of the most common objections to building richer models.
"They decided they were going to fly by instrument — not by what they could see."
Marketing Is a Nonlinear Multiplier — and Most Budget Conversations Ignore That Fundamental Asymmetry
The economic logic of marketing rests on a structural asymmetry that rarely surfaces explicitly in budget discussions. Sales performance is linear: doubling revenue targets requires, approximately, doubling the salesforce, because the output is the aggregate of individual quota attainment distributed across a bell curve. Marketing does not work that way. Because of the leverage already built into the function, growing revenue by a factor of two might require only a 20 to 25 percent increase in marketing spend — and in well-run B2B go-to-market operations, Stouse estimates the marketing-to-sales multiplier at somewhere between 10 and 20 times. The clearest test of this, he argues, is the counterfactual: remove marketing entirely, wait a year, and watch sales productivity collapse in ways the sales organisation cannot prevent because it has no mechanism to generate that leverage for itself.
The scaling question matters too. MMM is not exclusively an enterprise tool: smaller businesses, Stouse argues, use it primarily to manage cash-flow risk — modelling the downside of a bad spend decision before it drains liquidity. At the enterprise level, with marketing budgets running to hundreds of millions, the relevant frame is opportunity cost: every dollar allocated to marketing is competing against every other internal use of that dollar, and the CFO's question is not whether marketing works but whether it works better than the alternatives. That competitive framing, Stouse notes, is now effectively permanent — budget season has become a year-round condition.
"Marketing by definition is a nonlinear multiplier of areas of business performance that are linear — one of which is sales."
Summarised from RevOps FM · 47:45. All credit belongs to the original creators. Streamed.News summarises publicly available video content.
