Google Buys $5 Million of Toilet Paper a Year. The CFO Has No Idea.
Dale W. Harrison likes to run a poll: imagine a company about to sign a $5 million annual contract, renewed every year for five years. Who's on the buying committee? Most people guess the CEO, definitely the CFO, maybe a CTO. Then he reveals the deal: it's Google's contract for toilet paper in the employee restrooms. Google spends roughly $5 million a year on it. Nobody's CFO is in that meeting.
That poll is the setup for one of the sharper arguments Dale and Liam Moroney make in Episode 7 of B2B Effectiveness: Evidence-Based Marketing Ideas for B2B Practitioners: almost everything B2B marketers believe about who's actually in the room making buying decisions is fan fiction — and building targeting and lead scoring on top of that fan fiction guarantees failure before you've gathered a single data point.
Can better lead scoring define better targeting?
The episode opens with an inversion of the usual playbook. Most teams define an ICP, build targeting to reach it, then use lead scoring to check whether that targeting worked. Dale's argument runs the other direction: your own historical scoring data — who you've actually closed, and who you've actually lost — should be the thing that defines your targeting criteria, not the other way around. It's a feedback loop: day-one targeting with no data, gathering results, and using those results to progressively sharpen who you go after next.
Why lookalike lists keep pointing at your losses too
Dale walks through, mechanically, what happens when a marketer uploads a list of closed-won accounts to an ad platform to build a lookalike list. The platform breaks each account into descriptors — employee count, revenue, industry, geography — and goes looking for other companies matching those same descriptors. The problem is that most platforms only ever see the closed-won list. If 90% of those wins happened to be based in California, the lookalike list will skew hard toward California — even if 90% of the closed-lost deals were also based in California. A pattern that shows up equally in your wins and your losses isn't a pattern worth anything, no matter how confidently an algorithm reports it back to you.
Buying committees aren't new, and they're mostly invisible
A recurring theme in current B2B commentary is that “buying groups” are a new phenomenon marketers need fresh tools to handle. Dale's response: the term “buying committee” traces back to a Harvard Business Review article from 1965, describing research on large-company procurement that itself referenced practices going back into the 1940s. Nothing about how big organisations delegate purchasing decisions has changed in eighty years — what's changed is that vendors who've run out of credibility on one story need a new one to sell.
The committees themselves are also far more hidden than most targeting strategy assumes. Companies treat the existence of a specific buying group as confidential information — mention one by name on LinkedIn and, according to Dale, you can expect a call from HR within 24 hours, because HR actively monitors social media for exactly this. Dale's own experience sitting inside Fortune 500 procurement as a consultant backs this up with a smaller, funnier detail: in meeting after meeting, it wasn't a senior VP reviewing vendor materials, it was a summer intern handing out printed packets, because most committee members had never once visited a vendor's website.
The real de-anonymization numbers
This matters because most “account-based” targeting quietly assumes near-perfect visibility into who's actually involved. The real numbers, per Dale: roughly 65% de-anonymization is achievable at the account level, but only 5–7% at the level of a named individual. The rest is invention — a mental model built from imagination rather than evidence, which then gets treated as a targeting strategy.
Working backwards from your own history
The corrective is almost defiantly simple: look at who you've actually closed, and who you've actually lost, and find what's genuinely different between the two groups — not what's common to your wins in isolation, which tends to produce the same vowels-in-the-name nonsense the show has flagged before. Layer in buyer-seller fit (are you actually good at selling to companies that look like this one), and you get a targeting model built from your own evidence instead of a stranger's fan fiction.
How you'd actually know it's working
Dale's test for whether any of this is paying off is refreshingly concrete: track your MQL-to-SQL acceptance rate over time. If a five-minute qualifying call from a BDR reliably disqualifies 90%+ of what your scoring system hands over, the scoring system was never doing much work. The framing he returns to is a pair of slot machines — one paying off 1 in 100 pulls, the other 1 in 50. You don't need a dramatic improvement to double your results; given how bad most MQL-to-close rates already are (roughly 1%, per long-running Gartner benchmark data), even a small, honest improvement compounds fast.
Even in a hypothetical world where third-party intent data worked perfectly, Dale argues you'd still overshoot badly, because intent alone says nothing about whether a buyer is willing to buy from you specifically, or whether your sales team is actually equipped to close them. That's the same buyer-seller fit problem the series keeps circling back to, restated here through a sports analogy: a player who hits home runs for a living would be a disaster in the Tour de France. Being excellent at one thing doesn't imply competence at a related but different thing — and companies, like athletes, specialise.
Data is not information
The closing idea is one of Dale's most repeated, for good reason: data is not information. Knowing the last hundred outcomes on a roulette wheel tells you nothing about the next spin; knowing the last million tells you exactly as much. Most of what gets marketed as sophisticated, large-scale data infrastructure suffers from the identical problem — more volume, not more insight. The reassuring part, in Dale's telling, is that extracting the information that does exist rarely requires much more than the statistical sophistication of a motivated seventh grader with a spreadsheet.
Next week: how rising and falling markets change how many buyers are actually available to sell to — and why a category quietly being “eaten alive” can mean there's no one left in-market at all, no matter how good your targeting is.
