Free report: “5 Things $39.6M in B2B Programmatic Spend Taught Us.” Three years of managed spend and a billion tracked page loads, turned into five findings platforms won’t put in a pitch deck. Subscribe free to unlock it, plus every new issue of The B2B Stack.
Join 1,600+ B2B marketers, including readers at Dentsu, HubSpot, 6sense, Informa and DirecTV.
There’s a quiet revolution happening in B2B marketing.
If cookies were the currency of the performance era, HVAs are the currency of the Signal Era.
They bridge the gap between anonymous behaviour and actionable revenue. They turn marketing from a guessing game into a predictive model of demand.
This quiet revolution is less ‘adtech’, ‘martech’, or analytics dashboards. It’s a data currency
It’s the data trail buyers leave behind - the invisible moments that separate curiosity from commitment. The data unlock that far surpasses GA4, and traditional analytics in order to unlock the connection between accounts + signals + conversions - which transforms noisy data into a scientific way to understand purchase intent and map back to it with your marketing engine.
Most teams never see those moments. They lack the tech stack and data unlocks to make the invisible visible
As a result they chase clicks, other vanity metrics and fill reports with other weak proxy engagement metrics. This isn’t a knowledge or capability gap in these teams, there’s incredibly capable and smart people throughout B2B marketing, the vast majority of whom know these metrics are outdated and carry limited weight - its a technology and data gap. Its gaps like these that compelled me to start FunnelFuel and to get deep into solving these challenges.
There is a set of unlocks which can transform vanity into the scientific reverse engineering (or reverse mapping) of pipeline back to signals.
Cutting out and getting beneath the noise, there’s a smaller set of actions which present as repeating patterns that consistently appear before a deal closes.
TLDR - My Unlock: How I Turn High-Value Actions Into Pipeline
Most B2B marketers talk about High-Value Actions. Here’s how I take them further — connecting them to real pipeline outcomes.
Connect HVAs to closed-won deals – work backwards from CRM truth to see exactly what those companies did in the days, weeks, and months before they converted.
Map accounts to an Account Knowledge Graph – enrich every domain with firmographic, technographic, and operational context: what they do, where they do it, who they serve, and how big they are.
Cluster accounts by region, firmographics, and tech signals – because buying behaviour changes dramatically by market, company size, and industry vertical.
Reverse-engineer those clusters into on-site behaviours – use FunnelFuel Analytics to map real companies to their session data and reveal what “active intent” actually looks like across different clusters
Use ML/AI to identify the recurring pre-purchase behaviours – surface the patterns that consistently appear before deals close within each cluster.
Codify those patterns into HVA account-level buying signals – formalise them as first-party signals that define sales-ready activity.
Weight and score each signal by cluster – build intent models that adapt locally and by business type, producing a dynamic, first-party intent score for every account.
Pipeline those segments back to media buying – push live intent data into DSPs to adjust bid price, pacing, and creative sequencing toward the finish line.
Identify new prospects through early-stage HVA triggers – capture companies showing the earliest buying signals, map them to ICP clusters, and activate them first — owning the signal early to own share of voice early.
Lets dive in
What High-Value Actions Actually Are
HVAs are not vanity events.
They’re account-level buying signals - digital behaviours that correlate to companies who actually became customers.
Think of them as the connective tissue between your website analytics and your CRM truth.
Most platforms can tell you what happened. HVAs tell you what mattered.
Now we’re zooming in on signals that are statistically proven to matter as opposed to vanity metrics which represent proxies for intent - this is a great unlock as a starting point
The Core Idea: Reverse-Engineering Pipeline
But now we get onto the REAL unlock.
Instead of guessing what intent looks like, we work backwards from the companies that already bought.
We take the CRM’s converted accounts - the ones marked “closed-won” in Salesforce, Hubspot etc - and we trace what those companies did on your site in the days, weeks, and months before they purchased. We’re looking for patterns in the data, across signals like visit cadence, pages viewed, actions taken on site, content read, downloads, form fills etc.
We know these signals are scarcer because the modern buying journey happens increasingly with the vendor out of the loop BUT we’re working in the modern world and we need to adapt. This would have been even more powerful 5 years ago and it may be weaker in 5 years time - but for today, it remains incredibly powerful.
Having captured and joined the data, we use AI and machine learning to look for shared patterns:
What pages did they view?
How many visitors came from that domain?
How often did they return?
Which regions or industries showed similar behaviour?
That process is called reverse-engineering intent into pipeline.
It doesn’t just tell us that they bought — it reveals how they bought.
The process can be refined to break out vendor products lines too. Individual clusters of behaviours mapped to specific offerings to specific company types in specific locations can all be micro mapped. ML/AI crunching power makes this sort of granular slicing doable
Why We Segment by Region and Firmographics
Buying journeys aren’t universal.
A 500-person SaaS company in Berlin behaves differently from a 5,000-person manufacturer in Boston.
That’s why we split our models by:
Region – to respect local buying rhythms and content preferences, inclusing how well optimised for region the vendor may or may not be
Firmographics – to reflect company size, industry, and complexity of decision-making
This is how we build the most nuanced picture possible of what “sales-ready” looks like in every segment.
The Pattern That Changes Everything
Once the models are trained, certain behaviours jump out again and again.
For example, a recurring pattern might show that enterprise deals were often preceded by:
3+ unique visitors from the same account in 3 days
Each averaging 3+ high-intent pageviews per session
At least one visit to a pricing, integration, or documentation page
That cluster of actions becomes an HVA signature - a pattern that predicts purchase likelihood far more accurately than clicks or time-on-page ever could.
Below is the step-by-step method I use to reverse-map HVAs from won deals.

