CRO In A Highly-Regulated Industry Cautions What to Watch For When Doing An AI Overhaul
See how healthcare CRO John Stamatopoulos built an AI-powered revenue engine to improve targeting, forecasting, and sales execution in his highly regulated industry.
John Stamatopoulos
CRO of Brook Health
John Stamatopoulos is CRO of a Series-B remote-care company called Brook Health, where he is using AI to optimize nearly every function within the revenue pipeline.
We sat down with John to learn how he created a real-time signal engine that upended his entire workflow. He told us how it works, and shared the problems revenue leaders need to watch out for when integrating AI.
The Revenue Organization
My name is John Stamatopoulos, and I'm the Chief Revenue Officer at Brook Health, where we're focused on making continuous remote care and chronic care management accessible, scalable, and sustainable for health systems across the country.
Want more from The CRO Club?
Create a free account to continue reading and join modern revenue leaders using proven playbooks, peer insights, and AI-driven strategies to build predictable pipeline and scalable growth.
Have an account? Log In
Brook Health is a Series B growth-stage remote care company purpose-built for health systems and physician groups to operationalize chronic disease management at scale. Our revenue model is B2B enterprise. We contract directly with health systems, IDNs, and large physician practices. Our recurring revenue ties primarily to physician fee-for-service. We work incident-to-clinicians to help them drive revenue, and we have an associated fee schedule for the services we provide.
The revenue organization I lead spans sales, partnerships, and account management, with team members across the US. Our GTM structure centers on a health system sales motion. We have shortened cycles, but still navigate multiple stakeholders and a value proposition that must resonate simultaneously with the CFO, CMO, and the operations leader who implements the program.
Our model is interesting because we're not just selling software or devices. We're selling a clinical and operational capability. That means customer success isn't a post-sale support function; it's a co-deployment partner that drives the outcomes our contracts measure. Revenue retention and expansion live inside those clinical relationships, which fundamentally shape how I've built the team and how we think about pipeline, forecasting, and growth.
We've proven the model and are now scaling. This is exactly the inflection point where GTM discipline, AI augmentation, and cross-functional alignment matter most.
Building A Real-time Signal Engine
My path to this moment has been anything but linear, which makes it relevant. I spent the first half of my career in big pharma and device, and the last half launching digital health companies. I've built and rebuilt go-to-market teams across multiple remote care organizations, navigating CMS reimbursement shifts, value-based contracting cycles, and the perpetual complexity of selling into health systems — simultaneously, customers, partners, and competitors.
I've learned through all of this that revenue leadership in healthcare is fundamentally a data problem disguised as a relationship problem. The deals that stall don't stall because someone didn't like you; they stall because you couldn't connect your solution to the right clinical outcome, or financial trigger at the right moment in the right conversation. That's where AI changes everything.
Now, predictive forecasting isn't just a CRM dashboard; our tech stack is a real-time signal engine telling you which health system is nine months from a contract renewal, which CMO just published a population health initiative, and which clinical champion is losing executive air cover. When you layer that intelligence into how your sales, marketing, and account management teams operate, you stop running a revenue function and start running a revenue organism. Hopefully, one that learns, adapts, and compounds.
I didn't come up through a traditional SaaS playbook. I came up through the complexity of pharma, device, and digital healthcare delivery, where patient outcomes are at stake, and sales cycles span quarters and years, not weeks. That context explains why this AI moment matters so much to me: The teams that figure out how to augment humans with technical sophistication in highly regulated environments will define the next decade of healthcare commercialization.
The deals that stall don’t stall because someone didn’t like you; they stall because you couldn’t connect your solution to the right clinical outcome, or financial trigger at the right moment in the right conversation.
How AI Enhances The Entire Revenue Pipeline
A year ago, our pipeline reviews were intuition-driven. Reps gave verbal updates, and I triangulated between emails in the CRM, meeting notes, gut feel, and deal age to distinguish real opportunities from wishful thinking. In a complex health system sales motion with long cycles and multiple stakeholders, this is a dangerous way to run a revenue organization.
We built a connected sales tech stack to power every layer of revenue generation and management. It starts with Definitive Healthcare, which feeds market intelligence and health system data directly into HubSpot via a custom integration that populates deal-specific fields. Every account record reflects real-world context, not just what a rep manually typed after a call.
From there, we automatically enroll the account in an outreach sequence. If we have direct contact information, HubSpot runs the sequence. If we don't, LinkedIn Sales Navigator fills the gap, giving our reps a warm path to the right executive even when starting cold. Either way, the rep wakes up with a prioritized list of targets, pre-populated account context, and an outreach sequence already in motion.
What used to take a rep half a day of research and manual data entry now happens overnight. The rep's job shifts from building the foundation to having better conversations once the door opens. This workflow, in its simplest form, has meaningfully improved our efficiency in moving from market intelligence to active pipeline.
Additionally, we connected everything to Gmail and Zoom, and we use call intelligence features to monitor our teams' talk tracks and measure buyer sentiment in real time. That visibility changed how I coach and how quickly we can course-correct on a deal that's drifting.
In addition, we use Claude to automate and accelerate work that used to slow us down. Building ROI models, drafting case studies, assembling customized decks we can deploy quickly for a specific health system or IDN, and ensuring follow-up never falls through the cracks. What used to take days of prep now takes hours.
Everything downstream of pipeline creation improved. Forecast conversations became honest because the data was clean. Win rates improved on deals that previously would have gone quiet. Our team spends significantly less time on administrative work and significantly more time in front of buyers.
How AI Redefines Target Identification and Personalization
Before AI, identifying the right targets was manual and time-intensive. Reps spent hours cross-referencing data to identify health systems ready to move. Good targets were missed because reps lacked the hours to find them all.
Now, AI largely automates that process. Definitive Healthcare feeds our CRM health system and physician group data, filtered against our ideal customer profile. Patient population, chronic disease burden, technology infrastructure, geography — all are scored and prioritized without a rep lifting a finger. We surface opportunities that would have been invisible a year ago.
Personalization has changed just as dramatically. If a health system just announced a population health initiative or hired a new CMO, AI automatically pulls that context into the outreach. The message references something real and current in their world, rather than a generic value proposition that could apply to anyone.
Knowing exactly who to target and delivering genuinely relevant messages changes how you enter a conversation entirely.
How AI Integration Improves Revenue Outcomes
We have seen real wins with AI. Cycle times compressed meaningfully. When your team has clean data, clear account plans, buyer intent signals, and AI-assisted follow-up, deals move faster. We do not wait for reps to manually build an ROI model before a CFO meeting or remember to send a follow-up. That work happens automatically, keeping momentum alive between touches.
Lead generation also accelerated. Definitive Healthcare feeds our CRM and LinkedIn Sales Navigator fills in contact gaps, allowing us to identify, prioritize, and reach the right targets faster than we ever could manually. Top of funnel is cleaner, and conversion from prospect to active opportunity has improved.
Forecast accuracy is better, and I want to be clear that standardization, not just AI, contributes to this. We built a process requiring data hygiene and consistent stage progression. The AI reinforces that discipline rather than compensating for the absence of it.
The honest challenge is adoption. The tech stack only works when the team uses it consistently. Achieving this required coaching, reinforcement, and connecting tool usage directly to performance expectations. That cultural shift took longer than the technical implementation.
Why Humans Are Always In The Loop
So, AI informs or standardizes most operational mechanics of our revenue function. Pipeline prioritization, deal scoring, forecasting, territory planning, churn risk, activity tracking, and follow-up sequencing all run through our tech stack. AI surfaces signals, flags anomalies, and keeps the team accountable to process.
We made that a deliberate choice. In a scaling organization, the fastest way to lose revenue is to let operational discipline depend on individual habits. Standardizing and automating the mechanics creates a floor where everyone operates at a baseline of excellence regardless of tenure.
But humans still make decisions where trust, judgment, and relationship context are the actual product. Navigating a deal where the clinical champion and the CFO are misaligned. Knowing when to walk away, when to escalate, and when to restructure a contract to protect a long-term account. Pricing conversations that are contextual, relational, and politically sensitive inside a health system.
AI can model the ROI and build the case. Humans read the room. AI runs the system, and humans run the relationships. In healthcare, you cannot have one without the other.
Why AI Isn't Meeting Expectations in B2B Marketing
There are also places where AI hasn't performed as hoped.
As we leaned more heavily into AI-generated content and automated outreach, our impressions dropped. Health system executives are sophisticated buyers who have seen every templated sequence and AI-polished email. When content feels generated rather than genuine, engagement falls off. And to make matters worse, AI doesn't always have the context and can therefore generate technically accurate messages that are tonally wrong for the specific relationship. AI can produce content at volume, but volume is not the same as resonance.
The other gap is net new market development. AI is excellent at optimizing what is already in motion. But identifying genuinely new markets and crafting the narrative that opens a cold door still requires human creativity and market intuition that our tools have not replicated.
The broader lesson is that AI amplifies what is already working. Where the underlying strategy or message is weak, AI scales the weakness just as efficiently as it scales the strength.
Why Revenue Leaders Must Walk Through The Entire AI Workflow
You have to understand the downstream impact of efficiency gains in one area. When you automate and accelerate one part of the revenue process, you don't always realize you're creating more work elsewhere until it becomes a problem.
When we sped up lead generation and top-of-funnel activity, it created volume that our customer success and onboarding teams couldn't absorb. The tool worked exactly as intended. We just hadn't mapped what that success would demand from adjacent teams before we flipped the switch.
Before you deploy, walk through the full workflow and ask who else gets busier when this works.
What To Watch For As Your Team Adopts AI
Here's what you need to watch out for with your team:
- They may get out of the habit of reading what they send. The content looks polished, so people assume it is correct. In healthcare enterprise sales, where a single factual error or tone-deaf sentence in a message to a CMO can damage a relationship you spent months building, that assumption is dangerous. We built proofreading and review back into workflows as a non-negotiable step, not an afterthought.
- They may get overwhelmed and quietly stop using the system. It usually starts with information overload. The tools surface more data, more signals, and more tasks than the team is used to managing, and without clear instruction on where to start, people shut down rather than dig in. You can see it in CRM activity dropping off, unreviewed sequences, and reps reverting to their habits from before the tools existed.
- They may distrust the process. Some people are set in their ways, and no amount of dashboard access changes that. They will nod in the training and then go back to working their own system in parallel. The fix is not better technology. It is better onboarding. Start your team on one workflow, make it simple, make it win early, and build confidence before you add complexity.
Why The Deal Room Needs To Evolve
The deal room still needs to be redesigned. Specifically, what happens to a deal after heavy qualification. Most revenue organizations still treat this stage as highly manual. Reps coordinate follow-up, build materials, schedule stakeholders, and manage internal approvals largely on their own. AI can take over here, which is where most CROs are leaving efficiency on the table.
We are working toward an unstaffed virtual deal room. Once a deal reaches a defined stage, the system takes over operational mechanics. AI builds and updates the ROI model, generates customized materials for each stakeholder, manages follow-up sequencing, monitors engagement signals, and flags when human intervention is needed.
I have said throughout this conversation that humans need to be in the loop, and I stand by that. However, I want to nuance it. A well-trained video or a scripted walkthrough built around how a specific human thinks and communicates can act as a credible proxy for that person at key deal stages. You do not always need a live human in the room, but you need their perspective, judgment, and voice present. AI and video can carry that forward in ways not possible even two years ago.
The organizations that redesign their deal rooms around this thinking will not just be more efficient. They will be structurally harder to compete against.
Why Claude Cowork Is A Must-have
Claude Cowork is my must-have tool.
Its range makes it irreplaceable for me. It operates across every function, spanning operational, creative, strategic, and tactical. In a single session, I can build an ROI model, draft a deck, refine a contract narrative, and map out a go-to-market strategy. Nothing else in our stack does that.
Every other tool we use is purpose-built for one job. Claude connects all the jobs together. That flexibility makes it the one I would never give up. It doesn't always meet my expectations, but it still appears to do the most.
How To Budget For AI And Make Cross-Functional Decisions
Nobody talks openly about how to budget for AI and make decisions about tools that interact cross-functionally, yet it’s something every CRO wrestles with.
A few targeted tool investments quickly became expensive. When every team member finds a product they love and expenses it independently, you end up with a fragmented, costly, and often incompatible stack before anyone realizes what happened.
We decided to standardize, which was the right call. But even with standardization, we are still learning how to forecast costs accurately. Token usage, seat licenses, integration fees — AI tool billing models are not intuitive, and they scale in ways that are hard to predict until you are already over budget.
The cross-functional aspect adds another layer. Sometimes you cannot connect one team's tool to the broader AI infrastructure because doing so creates a compliance or security risk for another team. So you end up with islands of automation that cannot communicate, which undermines the connected revenue engine you were trying to build in the first place.
My advice is to build a cross-functional AI governance group early, set a centralized budget before individual teams start spending, and treat tool selection as an enterprise decision rather than a departmental one. We learned most of this the hard way.
What CROs Need To Know Before AI Revenue Initiatives
I wish I had known how many internal hoops I needed to jump through before a single tool went live.
We had a clear vision for what we were building. The business case was strong. I didn't fully anticipate how much approval I needed from IT, engineering, compliance, and cybersecurity before deployment. In healthcare, these are not bureaucratic formalities. They are legitimate gatekeepers protecting patient data, financial systems, and organizational liability. But the timeline these reviews add can quietly kill momentum and frustrate a team eager to move.
If I had known, I would have started those conversations at least two quarters earlier. I would have brought compliance and IT into the planning process before selecting vendors, not after. And I would have set expectations with my team upfront that the implementation timeline would be longer than the sales cycle for the tools themselves.
We could have avoided the scramble. Auditing systems for PHI exposure, establishing data boundaries, provisioning separate environments for tools unable to touch sensitive financial data, all of that work was real and necessary. It would have been far less disruptive if we had planned it rather than reacted.
In a regulated industry, your internal stakeholders are part of your go-to-market plan. Treat them that way from day one.
How Revenue Leaders Should Approach AI
The teams that win are not the ones with the most tools. They are the ones with the deepest discipline around the fewest — and right — tools.
My advice? Start small. Pick one thing and get really good at it before you add the next layer. We took on too much too fast, and some of it was not ready for prime time. The promise of a fully connected AI-powered revenue engine is real, but the path there is sequential, not simultaneous.
Prepare for AI to generate more work before it removes it. We had to audit and protect systems containing PHI, establish clear data boundaries, and in some cases provide team members with separate computers to keep sensitive financial data isolated, we wiil never take any chances with partner or patient data. This operational overhead is real, and most vendors will not mention it in the demo.
Once you find what works, lean in heavily. Build the habit, drive adoption, and let it compound before adding another tool. The teams that win are not the ones with the most tools. They are the ones with the deepest discipline around the fewest — and right — tools.
Finally, have realistic expectations. AI multiplies the force of a revenue strategy that already has sound fundamentals. If your pipeline is weak or your messaging is off, AI will surface those problems faster, but it will not solve them. The CROs who thrive are the ones who use AI to sharpen their thinking, not replace it.
Follow along
You can follow John Stamatopoulos' work on LinkedIn.
More expert interviews to come on The CRO Club!
