.png)
Take every account your company has ever closed-won. Upload the list to an ad platform and ask it to find more companies that look like them.
There's an obvious problem with that approach if you think about it literally. Every company on your closed-won list has vowels in its name. So, by the logic of a model looking for patterns in your winners, vowels are a characteristic associated with success. You could, in theory, use that information to find more companies with vowels in their names.
Obviously, that would be ridiculous. But it illustrates a real weakness in B2B lookalike modeling and target-account selection. Finding characteristics your successful customers share doesn't tell you whether those characteristics have anything to do with why they became customers.
That's where Episode 6 of B2B Effectiveness: Evidence-Based Marketing Ideas for B2B Practitioners picks up. In this solo episode, Dale W. Harrison, Chief Strategy Officer at The Insight Collective, builds on the previous episode's discussion of Buyer–Seller Fit to explain what a lead score should actually measure, introducing the Next-Gen MQL.
The MQL isn't broken. The scoring underneath it is.
The argument has run through the series from the beginning: the MQL serves a necessary function. Marketing can reach hundreds of thousands of people with a campaign, but a salesperson can only work through a limited number of conversations. Some form of qualification is needed to decide which prospects deserve that one-to-one attention.
The difficulty comes when that qualification is treated as a prediction it can't reliably make.
Long-running B2B benchmark data cited in the episode suggests that only around 5–6% of MQLs progress to a sales-qualified opportunity, with fewer ultimately becoming closed-won. If most leads passed to sales never become meaningful opportunities, the qualification process isn't doing a particularly good job of identifying where sales effort is likely to pay off.
Why finding patterns in your winners isn't enough
A typical lookalike model starts with a list of customers and identifies characteristics they share, such as industry, company size, geography or technology usage.
But a characteristic shared by your winners is only useful if it helps distinguish them from your losses. The vowels example illustrates how easily a model can identify patterns that have no meaningful relationship with commercial success.
The same issue affects conventional points-based scoring. A downloaded eBook might be worth ten points and a pricing-page visit two, but those values are often based on assumptions rather than a demonstrated relationship with the outcome. Adjusting the points can change the score without improving its predictive value.
The important question isn't whether a pattern exists. It's whether that pattern helps distinguish likely wins from losses.
Why more data doesn't necessarily mean more information
Dale illustrates this with a blood test analogy. A doctor doesn't need to take every drop of blood in your body to diagnose a condition. A small sample contains enough information to perform the test. Taking ten times as much blood wouldn't make the diagnosis ten times more accurate.
The same principle applies to lead scoring. Ten additional engagement signals aren't necessarily ten new pieces of useful information. A model can collect enormous amounts of behavioral data and still have little insight into whether someone is likely to buy.
AI-driven scoring can process far more signals than a traditional model, but more inputs and processing power won't necessarily improve predictions if the signals don't reliably distinguish buyers from non-buyers.
What should a lead score actually measure?
Rather than trying to establish whether someone is ready to buy based on their digital activity, Dale proposes focusing on factors that can be estimated using historical commercial evidence.
The Next-Gen MQL brings three of these together:
- Category expansion: Is the market growing or shrinking in a particular geography? Market growth affects how many buyers are entering the category and therefore the likelihood of finding active buyers.
- Buyer–Seller Fit: Is an account likely to consider a company like yours, and is your sales organization equipped to win? A good fit for the product category isn't necessarily a good fit for a particular seller.
- Firmographic customer lifetime value: If the account becomes a customer, what is it likely to be worth? Two prospects with similar win probabilities can have very different long-term commercial value.
Together, these factors shift the purpose of qualification. Rather than claiming to identify leads that are ready to buy, the model estimates where sales effort is most likely to generate a return, using information that can be supported by evidence.
Thinking in expected value
This connects to the idea of thinking in bets, which has run throughout the series. Marketing decisions are made with incomplete information. You can't know every conversation happening inside a prospective account or which competitor a buyer might ultimately choose.
A useful qualification system therefore needs to account for both the probability of success and the potential value of that success.
For example, if an opportunity has a 20% probability of closing and would generate $250,000 in customer lifetime value, its expected value is $50,000. That doesn't mean the company will earn $50,000 from the opportunity. It's a way of expressing its potential economic value while accounting for the likelihood of winning.
That provides a more meaningful basis for prioritizing sales effort than an arbitrary 100-point threshold. The score expresses the relative attractiveness of an opportunity without pretending to guarantee its outcome.
A glass box, not a black box
The final consideration is transparency.
A useful model should explain why an account received its score. If the score changes, you should be able to see which factors contributed and by how much. That makes it possible to test assumptions, challenge results and understand where the model performs well or poorly.
The same evidence can also be used in both directions. Rather than simply scoring individual leads, it can help identify the characteristics of customers a business is most likely to win, creating a more evidence-based ICP. Targeting and qualification become connected parts of the same process.
That's the thinking behind the Next-Gen MQL: a qualification model built around probability, commercial value and evidence, with enough transparency to explain how it reaches its conclusions.
But the same evidence can also be used to improve targeting. Rather than simply scoring the leads that come in, what if businesses could use what they've learned from past wins and losses to identify better prospects in the first place?
That's where Dale and Liam go next in Episode 7 of B2B Effectiveness, exploring how lead scoring and audience targeting can work together to improve the quality of opportunities reaching sales.
