.png)
For much of the last decade, intent data seemed like one of the most promising ideas in B2B marketing. Track what buyers are researching, identify accounts showing signs of interest and focus your effort where the demand appears to be.
Jamie Hendrie built a significant part of his business around that thinking. For several years, the results seemed to prove him right.
Then market conditions changed, the models stopped delivering the same results and Jamie found himself questioning how much of that success had come from the models in the first place.
Dale W. Harrison uses an analogy to describe the problem. If you're already in the rainy season, chanting to the thunder gods will appear remarkably effective. But when the drought arrives, the temptation isn't necessarily to question whether the chanting ever worked. It's to chant louder.
That experience forms the starting point for Episode 1 of Revenue Dynamics: Marketing, Sales, and the Economics of B2B Growth, a new series in which Jamie and Dale W. Harrison explore how to put evidence-based marketing effectiveness into practice.
In this opening episode, Jamie reflects on his journey from embracing intent data to questioning the assumptions behind it, and what ultimately changed his approach to marketing and sales investment.
When everything seemed to work
Jamie founded Inbox Insight, now The Insight Collective, in 2010 as an email newsletter business. The model was relatively straightforward: put useful content in front of decision-makers, understand who engaged with it and pass those contacts to customers.
As B2B marketers embraced lead nurturing, they needed a steady supply of opted-in contacts. Inbox Insight helped meet that demand.
Around 2016 and 2017, intent data began gaining traction in the UK, following its growth in the US. Similar companies were adopting it and scaling quickly, so Inbox Insight formed partnerships, tested intent signals against its own data and began using them to prioritize accounts.
For a while, the results looked extremely promising.
The market boom that made everything look successful
Between 2019 and 2022, B2B technology experienced extraordinary growth. Interest rates were close to zero, venture capital poured into SaaS businesses and valuations rewarded revenue growth.
When Covid disrupted in-person events, marketing budgets that would otherwise have gone toward conferences and exhibitions shifted into demand generation.
In Jamie's words, the business absolutely flew.
Conversions looked strong and the intent models appeared remarkably predictive. While the familiar 95:5 rule suggested that around 5% of a market would be actively buying at any given time, Jamie's teams were finding figures closer to 10%, 15% or even 20%.
It seemed like compelling evidence that the approach was working.
But as Dale explains, rapidly growing markets naturally bring more buyers into the category. The question was how much of that success came from the marketing strategy and how much came from the market itself.
What happened when the market turned?
At the end of 2021, Inbox Insight took private equity investment. The business was growing around 30% annually, ahead of its category, and met its EBITDA target during the first six months.
Then interest rates rose, investment conditions tightened and the technology market slowed.
Jamie began seeing the effects in outbound performance and pipeline data. For a company whose marketing strategy and campaign approach were also central to the services it sold, that was particularly concerning.
If the models became less predictive at precisely the moment market conditions changed, how much of their earlier success had actually come from the models?
Questioning the assumptions behind intent data
By mid-2022, the market had changed dramatically, and Jamie began looking more closely at the assumptions underpinning the company's approach.
During a visit to the US, he reviewed his marketing team's media plan. The campaigns themselves looked sound, but the Ideal Customer Profile (ICP) had expanded over time to include more accounts showing intent signals.
In doing so, the business was increasingly targeting areas where competitors might be better positioned to win.
On the flight home, Jamie began questioning the entire model. Was he trying to fix a system while still thinking within it?
As a CEO backed by private equity, the possibility was uncomfortable. Perhaps the business wasn't the Netflix of its category. Perhaps it was closer to Blockbuster, facing a fundamental shift in the market it had helped build.
He began seeking perspectives from outside the business, including one particularly outspoken skeptic of intent data: Dale.
The question that changed Jamie's thinking
Dale had experienced his own version of this problem years earlier.
In the early 2000s, he developed technology capable of identifying more than 90% of website visitors at an individual level, far beyond what today's de-anonymization tools typically achieve.
He believed that knowing exactly who visited a website and what they did would make it possible to predict when they were ready to buy.
But after building sophisticated statistical models and testing them against real outcomes, he found that the behavior of people who were buying often looked remarkably similar to that of people who weren't.
That experience led him to a question he repeatedly put to Jamie: Compared to what?
Jamie had spent years examining engagement activity among accounts that eventually became customers. There was plenty of activity to analyze, and much of it appeared consistent with the assumptions behind intent data.
So he asked his data engineering team to examine the same signals among accounts the business had lost.
The patterns looked remarkably similar.
It became difficult to argue that engagement was a reliable indicator of buying likelihood when much the same activity appeared among successful and unsuccessful opportunities.
Why accurate data can still lead to the wrong conclusion
Dale illustrates the issue with a pharmaceutical example.
Imagine a company announcing that 60% of patients recovered after taking its drug. That sounds impressive until you discover that 59% would have recovered without it.
The 60% figure is accurate. But without the comparison, it tells you very little about the treatment's actual effectiveness.
The same principle applies to marketing attribution. Spending $100,000 on branded search and counting the deals that followed doesn't establish how many of those deals the advertising actually created. Some buyers may already have been planning to purchase.
Incrementality testing offers one way to examine that distinction, helping marketers separate outcomes influenced by their activity from those that might have happened anyway.
For Jamie, this changed how he viewed intent data. He didn't conclude that it had no value, but that it was being used too early in the qualification process.
Intent might help break a tie between otherwise attractive opportunities. It shouldn't determine which opportunities are attractive in the first place.
Rethinking who is worth pursuing
The same issue applies to targeting.
Most lookalike and account-targeting tools start with a list of closed-won deals. But finding characteristics those accounts share doesn't establish whether those characteristics are associated with winning. You need to compare them with the accounts you lost.
Jamie also recognized that winning an account doesn't necessarily make it a good customer.
A large contract might consume months of effort from sales, legal and procurement teams, only for the customer to leave after a year. If that account generated little economic return, finding more customers like it could make the business worse off.
Category fit alone wasn't enough, either. Two companies selling similar products could have very different chances of winning the same account, depending on their reputation, experience and capabilities.
This led Jamie toward Buyer–Seller Fit: understanding which customers a business is well positioned to win and serve profitably, rather than simply which organizations could buy its product.
Marketing as a capital allocation decision
For Jamie, the bigger shift was recognizing that data is only useful if it changes a decision.
When resources are finite, every choice involves a trade-off. Should the next dollar of marketing budget go into one audience or another? Should a salesperson spend the next hour pursuing this account or that one?
Jamie and Dale began applying ideas from investment theory, including portfolio management, to marketing and sales decisions.
Rather than optimizing isolated campaign metrics, the aim was to identify where the business could generate the greatest commercial return.
Looking at the economics of each opportunity
That meant bringing several factors together:
- Probability of winning: How likely is the business to win this account?
- Customer value: What might the relationship be worth over its lifetime?
- Sales effort: How much time and resource will it take to win and serve the customer?
- Market opportunity: How much demand exists, and is the category growing or shrinking?
These factors can lead to very different decisions from those suggested by conventional lead scoring.
A segment with a low close rate could still be commercially attractive if deals are valuable and relatively efficient to win. An enterprise account might look impressive but require more sales investment than the likely return justifies.
For Jamie, the ICP could therefore become something more meaningful than a static targeting document: an investment strategy that helps Marketing, Sales and Finance agree where the business should allocate its resources.
Building a glass box, not a black box
The approach Jamie and Dale have developed is built around a company's own historical sales data: the accounts it won and lost, the effort invested and the value those customers ultimately generated.
By examining how outcomes vary across different customer characteristics, the model identifies where sales effort is most likely to produce a worthwhile return. It can then inform ICP development, target-account selection, lead qualification and sales resource allocation.
Crucially, the calculations are transparent. Rather than relying on an unexplained score, businesses can see the evidence behind each recommendation and understand the trade-offs between growth and efficiency.
That's the thinking behind the Next-Gen MQL: bringing Marketing, Sales and Finance together around a more evidence-based approach to deciding where commercial investment is most likely to pay off.
