Twenty years ago, Dale W. Harrison built a system that de-anonymized roughly 90% of website traffic down to a named individual, then checked his scoring model against what actually happened next. The result: each of his two buckets — “likely to buy” and “not likely to buy” — was only about 60% correct. A small, genuine improvement over a coin flip, not the confident precision the intent data industry has been selling ever since.

In Episode 10, Diego Sosa joins Dale again for the most detailed breakdown yet of what should actually replace the broken MQL. They cover why you don't need to know why something works, only that it reliably does; the two-sided idea of buyer-seller fit; why thinking in slot machines beats thinking in vending machines; and why the data you need has been sitting in your CRM the whole time.

Timestamps

0:00 Cold open: intent data vendors are lying to you

0:20 Welcome back

0:33 Setting up today's topic: what replaces the old lead scoring model

2:53 20 years, still a 99%+ failure rate

5:34 It's just a slot machine, not a vending machine

6:07 Causality without correlation: the biotech and drug analogy

6:47 Viagra vs. antidepressants: two kinds of “we don't know why it works”

9:16 Defining buyer-seller fit

9:50 HubSpot vs. Salesforce in the Fortune 1000

10:39 Why Exxon won't buy from a six-month-old company

11:13 Seller fit: the biotech researcher-vs-government-lab example

17:26 Thinking in slot machines, not vending machines

19:32 The ROAS story, and “this is where marketers get stupid”

21:55 Why win rate and sell cycle are both the wrong things to optimise

23:27 Rolling buyer fit and seller fit into one metric: revenue per unit of sales effort

44:06 The Pampers analogy: reverse-engineering demographic targeting

47:26 Why the data you need is already sitting in your CRM

48:15 The vowels-in-the-name story, one more time

50:35 Diego's first sales job, and the deals that looked too easy

52:26 The notion of opportunity cost

55:48 Dale's own 20-year-old experiment: de-anonymizing 90% of traffic

59:47 Click monkeys, and the people who show up ready to buy with no warning

1:01:00 Naming it: Dark Social

Key Topics Discussed

- Why you don't need to know why something works, only that it reliably correlates with the outcome

- Buyer-seller fit: why it's not enough that a buyer would buy from you — your sales team has to be equipped to close them too

- The HubSpot vs. Salesforce Fortune 1000 gap, and why tenure and trust matter more than product quality

- Thinking in slot machines: why win rate and payout size can't be judged in isolation

- Why win rate and sell-cycle length are both the wrong things to optimise for

- Revenue per unit of sales effort: the one metric that rolls buyer fit and seller fit together

- Reverse-engineering targeting from your own CRM data instead of buying generic intent scores

- Click monkeys and dark social: the two failure patterns any real scoring model has to account for

Notable Quotes

“What we're really looking at here are slot machines, not vending machines.” — Dale W. Harrison

“You cannot work backwards from success. You have to look at how the winners are different from the losers, not how the winners are similar to one another.” — Dale W. Harrison

“Three minutes on the phone with someone will give you more information than 1,000 touch points from your intent data provider.” — Dale W. Harrison

Resources & Mentions

- Gartner B2B sales funnel benchmark data

- HubSpot and Salesforce Fortune 1000 market share data

Next Episode

Back next week with more detail on the buyer-seller fit model.

Subscribe & Follow

Catch every episode of B2B Effectiveness: Evidence-Based Marketing Ideas for B2B Practitioners on your favourite podcast platform, and subscribe here on YouTube for future episodes.

#B2BMarketing #DemandGeneration #LeadScoring #MarketingAnalytics