This week’s post is a little different. For the first time in our history, we’re sharing a behind-the-scenes look at some of the internal tools we’ve acquired & built at Pricepoint Partners to help B2B SaaS companies grow. 

I know not everyone can work with us directly, whether due to not having a referral, budget, or the wrong timing. So I’m giving access to this tool to 5 readers who either:

1. Book a quick call below, or

2. Email me with how questions/features/campaigns they’d want to test with it

For the majority of people I can’t give access to feel, free to copy the approach & build your own version. And if you do, I’m happy to help privately.

Book your free B2B SaaS strategy session. No sales pitch. 

• Need sharper GTM strategy? I’ll assess your current growth engine and find opportunities to unlock new revenue.

• Looking to improve efficiency? I’ll show you how top startups are using AI to reduce costs.

• Unsure how to leverage AI? Learn how the best in the world use AI to accelerate revenue growth.

👉 Book your free growth session 

I Scraped 500 CFO Profiles to Predict Their Buying Behaviour

Imagine trying to get 30 minutes with a Fortune 500 CFO to validate your product messaging. You'll send hundreds of emails, wait weeks for a response, and if you're lucky enough to get on their calendar, you get one shot to ask the right questions.

Meanwhile, everything you need to understand how they think, what drives their decisions, and what messaging will resonate is already sitting there in public, scattered across LinkedIn posts, conference talks, podcast appearances, and hiring decisions they've made.

The information is already public. You just need to know how to find it.

My team and I recently spent three months building "synthetic customer research”, using AI to create detailed personality profiles of target customers based on their digital footprints. The goal wasn’t to simply prepare better for conversations with decision-makers but replace them. In B2B, those conversations are expensive, laborious, and increasingly rare. Now, teams can access simulated buyer profiles anytime, and ask them unlimited questions.

This project was one example of how we’re building tools that deeply integrate AI into our consulting workflows, from marketing strategy to sales optimisation to product design. Scraping 500 CFO profiles and predicting how they'd respond to different messaging was a test case, and the results were incredible..

Facebook’s 2015 Study

This whole experiment started with a 2015 research paper from Cambridge and Stanford that I couldn't stop thinking about. Researchers analysed 86,000 Facebook users and found something remarkable: algorithms could predict personality traits more accurately than the people closest to them.

With just a few Facebook "likes," the algo could predict someone's personality better than colleagues, friends and even spouses:

• 10 likes: more accurate than a work colleague

• 70 likes: more accurate than a friend

• 150 likes: more accurate than a parent or sibling

• 300 likes: more accurate than a spouse

That was ten years ago. AI/ML has advanced dramatically since then.

If algorithms could understand consumer personality from social media activity in 2015, what could they learn about B2B decision-makers from their professional digital presence in 2025?

The 2023 AI agent simulation paper showed that AI agents could exhibit remarkably human-like behaviors when trained on structured context. With just one two-hour interview, researchers were able to simulate accurate responses from that person with 85% fidelity.

Building the CFO Database

We started with a simple hypothesis: execs leave enough digital breadcrumbs to predict how they'll respond to different business propositions.

We worked with one of our clients to build the CFO list using a combination of LinkedIn Sales Navigator, company websites, and public speaking directories. The goal was to focus on mid-market companies ($50M–$500M revenue) and gather enough context to understand real buyer behavior.

For each CFO, we collected:

• Professional Content: Every LinkedIn post, comment, and article they'd shared in the past two years. Conference presentations and podcast appearances where they discussed their priorities and challenges.

• Hiring Patterns: Recent job postings from their companies, especially finance and operations roles. What you're hiring for reveals what you're struggling with.

• Background Signals: Previous companies, career trajectory, and educational background. A CFO who came up through consulting thinks differently than one who grew through operational roles.

• Current Priorities: What they're posting about, what events they're attending, what content they're engaging with. Their digital activity reveals their current focus areas.

• Internal Data: Email sequences, sales calls, LinkedIn outreach, win/loss notes, call transcripts. All of it can be parsed and analyzed to reveal the messaging and angles that drive action.

The Personality Profiling Framework

Raw data is just noise without a framework to interpret it. I used the "Big Five" personality model that the Facebook study validated—the same framework psychologists use to measure personality traits:

• Openness: How receptive are they to new ideas and innovative solutions?

• Conscientiousness: How detail-oriented and process-focused are they?

• Extraversion: Do they prefer collaborative or individual decision-making?

• Agreeableness: How much do they value consensus and relationship preservation?

• Neuroticism: How do they handle risk and uncertainty?

But we added B2B-specific dimensions:

• Risk Tolerance

• Growth vs. Efficiency Focus

• Decision-Making Style

• Communication Preferences

Here’s an example profile that emerged:

Sarah, CFO of a $60M ARR SaaS company. High conscientiousness (mentions "process improvement" in multiple LinkedIn posts), low neuroticism (took CFO role at an early-stage company, frequently posts about "calculated risks"), high openness (regularly shares articles about AI and automation). Risk tolerance: Medium-high (based on previous career moves from Big 4 to startup to scale-up). Primary language: ROI-focused with frequent mentions of "efficiency" and "scalable processes."

What emerged from this process were personas our client could chat with based on real data from their ICP.

Testing the Predictions

As AI lowers the barrier to launching new products and features, the hard part is no longer building but differentiating.

Everyone’s saying the same things. Figuring out what angle works best in a cold email, billboard or even what design is most user-friendly is now critical in growth and survival.

We tested this with three consulting clients who were struggling with B2B messaging, product positioning, and go-to-market angles. Each had built second products without really understanding how their target customers would respond.

For example:

• A security compliance startup had two core ICPs: founders and COOs. Founders responded best to messaging about revenue opportunities unlocked by compliance (e.g., closing enterprise deals). COOs responded better to failure-prevention messaging (e.g., avoiding breaches or audit failures).

• A financial planning platform tested different language around automation and control. The version that emphasized freeing up headspace (“Automate the repetitive so your team can focus on strategy”) dramatically outperformed.

The results were striking. Claude 3.7 Sonnet, trained on public data correctly predicted message preference with 87% accuracy. Using personality-matched messaging improved response rates by 45% on average.

What This Means for B2B Marketing

This experiment convinced me it’s archaic not to test messaging first.

Most companies wouldn’t dream of launching a product without first testing it for bugs. Yet they’ll routinely launch six-figure marketing campaigns based on gut feel.

We think the same testing mindset should apply to messaging, new ads, designing new features, and go-to-market.

The Speed and Cost Revolution

Traditional approach:

• 3–6 months to recruit CFOs, schedule interviews, conduct research, and analyze results

• Cost: $30K–$100K for meaningful sample sizes

Synthetic approach:

• 2 weeks to build profiles, test messaging, and get directionally accurate predictions

• Cost: Under $100 in token costs

The quality isn’t perfect. It doesn’t need to be.
It just needs to be accurate enough to beat blind bets.

My Offer: Book your free B2B SaaS strategy session. No sales pitch. 5 Readers Get Free Access.

• Need sharper GTM strategy? I’ll assess your current growth engine and find opportunities to unlock new revenue.

• Looking to improve efficiency? I’ll show you how top startups are using AI to reduce costs.

• Unsure how to leverage AI? Learn how the best in the world use AI to accelerate revenue growth.

👉 Book your free growth session