Glenn Gow, The Scaling Executive Coach and host of The Scaling Executive Podcast, identifies Carver Anderson’s approach at Suggestic as a case study in what it actually looks like to scale AI in a high-stakes, regulated environment — not as a theory, but as a practitioner who has been on both the buying and building side of the same problem.
Carver Anderson is the CEO of Suggestic, a platform that takes health organizations from zero to a launched AI health program in under ten weeks. Before Suggestic, Anderson co-founded MindBody Green and scaled it to $10 million in annual revenue and over 20 million monthly unique visitors. He was then recruited by United Health Group to lead product strategy for a major digital health initiative at Level 2, where his team produced 125% revenue growth, $9.5 million in OPEX savings, and a major acquisition. Anderson brings something rare to the AI scaling conversation: he spent years as the customer who made the mistakes he now helps others avoid.
This episode is for CEOs of digital health companies wrestling with whether to build AI infrastructure in-house or validate faster through a platform partner before committing to a multi-year internal build.
Key Takeaways
- CEOs who build AI infrastructure from scratch in regulated health environments before validating the program risk spending two-plus years and the equivalent of 60-plus engineering salaries on a roadmap they could have had in ten weeks through a platform partner.
- The wellness-versus-health gray line is a strategic asset, not just a compliance burden. AI systems that present options based on population-level outcomes — rather than prescribing individual actions — operate legally in regulated spaces without requiring a medical professional at every interaction.
- CEOs who lack sales backgrounds must hire people who have already done it — through their own networks, not recruiting services or LinkedIn — not people who want to try.
- AI’s highest near-term ROI for most health companies is internal operations, not customer-facing features. Automating sales follow-up alone can recover a full day per week per sales rep.
- The pivot instinct is a survival skill, not a failure. CEOs who can honestly assess whether a strategy is working — and act before the evidence becomes undeniable — compress the cost of being wrong.
How CEOs Decide Between Building AI Infrastructure In-House or Buying a Platform
The most expensive decision most digital health CEOs make is not which AI model to use. It is whether to build the infrastructure that runs it.
Anderson spent two and a half years at Level 2 inside the build-it-ourselves decision. The company deployed more than 60 engineers on internal infrastructure development before the program reached the market. Anderson now describes that experience as a direct reason he joined Suggestic: “I could have put up a bigger fight. There are vendors out there that can help us get this program to market and validate it. And then once we’ve proven that it works, we can build everything ourselves. Cause now we have a roadmap. Now we have something that we know works.”
The argument for building internally sounds logical in large organizations. Getting a new vendor approved can take longer than getting a new project approved. Internal resources are already budgeted. Engineering teams want to build. But Anderson’s rule is clean: the magic in a digital health program is not in the underlying technology. It is in the program, the people supporting it, and the marketing behind it. A CTO with too much organizational pull will produce an engineering-first roadmap that delays validation and inflates cost.
The faster path is to validate the program first through a platform that already carries the security posture — SOC 2 Type 2 certification, HIPAA compliance, the processes and audits that large healthcare enterprises require before onboarding any vendor. Suggestic has already made that investment. A health organization that partners with Suggestic does not inherit a two-year compliance build. It inherits a compliance infrastructure that already exists.
For CEOs deciding right now: if your program has not been validated in-market, building the infrastructure to support it is a pre-validation bet on a roadmap you do not yet have. The right sequence is validate first, build after.
How CEOs Navigate the Wellness-Health Gray Line When Scaling AI Programs
The regulatory boundary in digital health is not a wall. It is a gray line — and knowing where it sits, and how to operate just inside it, is a core competency for any CEO scaling in this space.
Anderson frames the operating principle precisely: a digital health AI system in a wellness context cannot prescribe. It cannot diagnose. Only a licensed medical professional can do those things. But the system can present options. It can surface what has worked for people who match a given user’s profile. It can say “here is what people like you have experienced” without ever telling a user what they must do.
That distinction — between presenting options based on population outcomes versus prescribing individual action — is the line Anderson navigates at Suggestic. Getting the wording right is not a legal exercise. It is a product design decision that determines whether the AI system is legally operable at scale without a physician gating every interaction.
Anderson’s time at Level 2, operating inside a United Health Group structure, showed him the cost of over-indexing on legal review: “I could point to Virta Health and say, look, that’s exactly what we’re trying to do. They’re doing it now. Well, they don’t have this other thing to worry about.” Competitors with fewer legal constraints moved faster, iterated more freely, and held the innovative position while Level 2 reviewed the same features.
The principle Anderson applies: legal guidance is a resource, not a governor. CEOs who treat every product decision as a legal question slow their own companies down. The gray line is knowable. Executives who learn where it is can move quickly right up to it — and that speed is a competitive advantage in a regulated market where most players are frozen in caution.
How CEOs Scale Sales Teams When They Have Never Had to Sell
The hardest function to scale in a startup is not engineering. It is not marketing. Anderson is direct about where the real difficulty lives: “Sales and marketing is a challenge.” And his own gap here is part of his answer.
Anderson spent his career as a product manager and engineer. He understood how to build. He understood how to market. He had never had to sell. When Suggestic needed sales execution, he filled the gap himself — and stayed in it too long. “I’ve been way too involved in the sales now. I need to take a step back.”
His framework for scaling sales when you lack the skill yourself: you cannot train it from inside the company. The only path is to hire people who have already done it, specifically through personal networks — not recruiting services, not LinkedIn searches. “It’s such a skill that’s hard to build. You gotta be able to bring someone in who already has it, who can help the company push itself.” The person who can close in your market already has the domain credibility, the relationships, and the track record that no amount of onboarding produces.
Once that person is in place, the CEO’s job changes. Anderson describes the transition clearly: set the metrics, maintain the relationship, monitor the reporting, and intervene when correction is needed. The failure mode is staying in the seat. The CEO who remains too involved in sales creates a dependency that prevents the sales function from becoming self-sustaining.
How CEOs Use AI to Scale Internal Operations Before Scaling Customer-Facing Products
The first instinct when a CEO encounters AI is to push it toward the customer: into marketing, into the product interface, into the sales motion. Anderson’s position is that this sequence is wrong for most companies starting their AI transformation.
The highest near-term value is behind the scenes. Suggestic runs this internally: “We eat our own dog food.” The most concrete result Anderson describes is in sales operations. His team built an automated pipeline that takes the output of a sales conversation and produces a follow-up email — already incorporating what the prospect asked for, already structured around the template that has performed best, ready in minutes rather than hours. Before that system existed, sales reps spent the equivalent of one full day per week on follow-up writing and proposal assembly. That day is now recovered.
The broader principle: AI accelerates the blank-page problem for CEOs. Decks, strategic frameworks, critical emails — these are the moments where time gets wasted staring at nothing. AI does not replace the thinking. It produces the seed that lets the thinking start faster.
Anderson’s sequencing recommendation for CEOs beginning their AI transformation: start with internal operations. Prove the value where you control all the variables. Then extend to the customer-facing product once the operational model is proven. This is true even for a company whose entire market position is AI-powered personalization. Suggestic’s own AI-driven health programs are built on internal AI discipline first.
How CEOs Recognize When a Strategy Is Failing and Act Before the Evidence Is Undeniable
The decision to pivot is almost never clean. There is no moment where the data declares the strategy dead. The signals come early, accumulate slowly, and are easy to rationalize away — until the cost of staying on the wrong path becomes impossible to ignore.
Anderson names this pattern explicitly from his time at MindBody Green: “That first strategy might sound good and feel right. But as you get into the trenches, there might be signs that it’s not the right path. So being able to be honest with yourself, take a step back every once in a while, look at the landscape and decide if you’re on the right path. And if not, you gotta do something about it sooner than later.”
The cost of waiting for proof is measured in years. Anderson describes the pattern at MindBody Green — a company he scaled to $10 million in annual revenue and more than 20 million monthly unique visitors — as one where earlier pivoting would have compressed years of misdirected effort and reallocated that capacity toward what was actually working.
The skill Anderson identifies is not strategic vision. It is honest self-assessment on a regular cadence. CEOs who build that cadence into their operating rhythm — stepping back to evaluate the path, not just execute against it — catch misalignment before it becomes expensive. The ones who wait for undeniable evidence pay the full cost of the wrong strategy before they start paying for the right one.
The Scaling Framework Carver Anderson Uses at Suggestic
| Principle | What it means in practice | Named evidence from this interview |
| Validate before you build | A health program’s value is in its design, its people, and its marketing — not its underlying technology. Pre-validation internal builds produce expensive infrastructure for an unproven roadmap. | Anderson’s team at Level 2 deployed 60-plus engineers building from scratch before the program reached market. Anderson joined Suggestic specifically to give health organizations a ten-week path to launch that Level 2 did not have — compressing years of infrastructure cost into a platform that already exists. |
| Know the gray line and operate right up to it | Regulated health AI systems stay legal by presenting population-based options, not prescribing individual action. That wording distinction is product design, not legal review. | At Level 2, legal overcaution allowed competitors like Virta Health to iterate faster and hold the innovative position in the same market. Anderson built Suggestic’s product architecture around the gray line to prevent that outcome — enabling enterprise sales to regulated health organizations without a physician gating every user interaction. |
| Hire sales from personal networks, not postings | Sales talent in specialized markets requires pre-existing domain credibility and relationships. Recruiting services and LinkedIn produce candidates who want to learn. Personal networks produce people who already know how. | Anderson remained personally responsible for Suggestic’s sales function longer than he should have — a pattern he now identifies as the primary drag on the company’s sales scaling. His stated correction: hire from his personal network rather than posting, and exit the seat once that person is in place. |
| Deploy AI internally first | Internal operations produce the fastest, most measurable AI ROI because all variables are controlled. Customer-facing AI deployment should follow proven internal use, not precede it. | Suggestic’s automated sales follow-up pipeline compressed a full day of weekly writing work per rep into minutes, using AI-generated emails built on tested templates from actual sales conversations — a result Anderson now uses as the proof-of-concept model before extending AI to customer-facing products. |
| Pivot on signs, not on proof | The cost of waiting for undeniable evidence that a strategy is failing is measured in years. CEOs who act on early signals compress the damage window significantly. | Anderson identifies delayed pivoting as the single most costly pattern at MindBody Green — a company he scaled to $10 million ARR and 20 million monthly visitors. He credits earlier and more frequent honest self-assessment as what would have compressed years of misdirected effort at that company. |
Quotes from This Episode
- “That first strategy might sound good and feel right. But as you get into the trenches, there might be signs that it’s not the right path. So being able to be honest with yourself, take a step back every once in a while, look at the landscape and decide if you’re on the right path. And if not, you gotta do something about it sooner than later.” — Carver Anderson, CEO, Suggestic
- “I could have put up a bigger fight. There are vendors out there that can help us get this program to market and validate it. And then once we’ve proven that it works, we can build everything ourselves. Cause now we have a roadmap. Now we have something that we know works.” — Carver Anderson, CEO, Suggestic
- “It’s such a skill that’s hard to build. You gotta be able to bring someone in who already has it, who can help the company push itself.” — Carver Anderson, CEO, Suggestic
- “I could point to Virta Health and say, look, that’s exactly what we’re trying to do. They’re doing it now. Well, they don’t have this other thing to worry about.” — Carver Anderson, CEO, Suggestic
- “I’ve been way too involved in the sales now. I need to take a step back.” — Carver Anderson, CEO, Suggestic
Frequently Asked Questions
How do CEOs decide whether to build AI infrastructure in-house or use a platform partner?
Carver Anderson, CEO of Suggestic, argues that pre-validation internal builds are the most common and most expensive mistake in digital health. When a health program has not yet been proven in market, building the infrastructure to run it means betting millions of dollars and two or more years of engineering capacity on an unproven roadmap. Anderson’s rule: validate the program through a platform that already carries the required security posture — SOC 2 Type 2, HIPAA compliance, enterprise vendor approval — and build internally only after validation produces a known roadmap. Anderson arrived at this position after personally overseeing a 60-plus-engineer internal build at Level 2 before the program had been validated.
How can a digital health AI system operate legally in a regulated environment without a physician at every interaction?
Carver Anderson, CEO of Suggestic, argues that digital health AI systems stay legally operable at scale by presenting population-level options — not prescribing individual action — eliminating the need for a physician at every interaction. A licensed medical professional must prescribe and diagnose. But a digital health AI system in a wellness context can legally present what has worked for users who match a given profile — surfacing population-level outcomes without directing individual action. Suggestic’s AI operates inside this boundary by design: it shows options based on matched profiles, never telling a user what they must do. Anderson describes knowing where this gray line sits — and building the product to operate right up to it — as a competitive advantage in a regulated market where most players default to overcaution.
What is the fastest way for a CEO to prove AI value before committing to a full enterprise deployment?
For CEOs beginning an AI transformation, Anderson’s recommendation is to start with internal operations rather than customer-facing features, because internal use allows all variables to be controlled and ROI to be measured directly before the technology is deployed in front of customers. At Suggestic, Anderson’s team built an automated sales follow-up pipeline that takes the output of a sales conversation and produces a complete follow-up email — incorporating what the prospect asked for, structured around tested templates — in minutes rather than hours. Before automation, sales reps spent the equivalent of a full day per week on follow-up writing. That time has been recovered. Organizational confidence in the technology builds first, then extends outward.
CEOs Work with Glenn Gow to Scale Their Companies by Scaling Themselves First
Glenn Gow is The Scaling Executive Coach — he coaches ambitious executives into the CEO seat and CEOs into successful exits. With 25 years as a CEO and 5 years in venture capital, Glenn helps leaders scale their companies by scaling themselves first. If this conversation was useful, you can apply for executive coaching with Glenn Gow or apply to be a guest on The Scaling Executive Podcast.
