CPO Rising Series: BetterHelp CPO on Why CPOs Need to Think Like Business Operators

What does it mean for a CPO to truly own outcomes? In this episode of the CPO Rising series hosted by Products That Count Resident CPO Renee Niemi, BetterHelp CPO Terre Layton will be speaking on how product leadership is evolving in the age of AI, drawing on a career that started with AI research long before LLMs existed. Terre shares why CPOs need to think like business operators, how she thinks about building teams for the agentic era, and why time to value and retention are the two metrics that tell her everything.

Join us for new conversations with leading product executives every week. Roll through the highlights of this week’s event below, then head on over to our Events page to see which product leaders will be joining us next week.

Show Notes:

1. Terre discovered product management by accident after realizing she was already doing the most interesting parts of the job as a human factors engineer. The moment she was given the opportunity to be both PM and designer on a new product, she knew she’d found her calling.
2. The CPO role has shifted significantly in the last five years from roughly half operational to predominantly strategic. The biggest change is that CPOs are now expected to be business operators who own outcomes directly, with product as their primary lever.
3. AI strategy has moved from the CTO’s domain into the CPO’s. Because AI now spans product surface, cost center, and capability layer simultaneously, it has to live in product for the strategy to be coherent.
4. Terre has been working with AI since graduate school, initially in linguistic parsing and natural language processing long before the LLM era. She says that while she’s seen multiple waves of AI development, today is genuinely unprecedented.
5. Running product in a public company is fundamentally different from a private one. In a public company, predictability matters more than vision and storytelling, accountability cycles are much shorter, and every decision gets filtered through quarterly investor expectations and risk management.
6. The next evolution in experimentation will shift from stage-gated processes to something continuous and probabilistic. When AI agents can surface signals across disparate data sources and cohorts, product teams will be able to make resource decisions backed by probabilistic outcome estimates rather than gut feel.
7. Terre protects a dedicated generative and agentic AI pod even when the rest of the business is under pressure. Having a team ring-fenced from day-to-day urgency is how she ensures the long-term bets that produce step-change outcomes don’t get sacrificed for short-term fires.
8. A new function is coming to product teams: people responsible for managing and QA-ing agents. Just as product QA monitors production behavior that’s hard to test in staging, someone will need to ensure agents are consistently meeting their defined expectations in the real world.
9. Prompt engineering is not a dedicated role; it’s a skill everyone on the product team needs. With tools like Claude Code, Cursor, and Lovable, you can’t use them effectively without understanding how to prompt well.
10. Having a great CPO is like setting the highest order bit to one: when product-market fit is strong and a great product leader is in place, everything else in the business gets easier. Financials, marketing, talent retention, and scaling all compound from that foundation.
11. Great product leaders don’t just identify where to invest resources; they also recognize product expansion opportunities that act as a flywheel for growth. That combination of capital allocation and growth identification is what creates lasting investor returns.
12. Terre’s two most important metrics are time to value and retention. Getting users to the “aha moment” quickly signals that something real is working, and retention is what determines whether the business is fundamentally healthy or quietly in trouble.
13. BetterHelp’s matching algorithm is one of Terre’s biggest AI wins. A major consumer study confirmed that fit is the number one thing clients care about when seeking therapy, and BetterHelp’s algorithm delivers it better than anyone else in the space.
14. Using Cursor AI, BetterHelp’s engineering team went from processing one email in five hours to five emails per hour during a complex localization project. Across 350 emails, that efficiency gain was the difference between an impossibly long slog and a feasible launch timeline.
15. BetterHelp launched five new international markets in 2025, a major undertaking given the platform’s technical debt. Terre describes trying to localize a product as the single fastest way to discover exactly how much technical debt you actually have.
16. Terre organized her product org around four clear functions: product management, product design, product research, and product ops. That structural clarity gives her team the ability to operate with autonomy at scale.
17. AI won’t replace product managers, but product managers who are complacent and don’t embrace AI will be replaced by those who do. It’s not an AI-versus-human story; it’s a story about which humans understand how to use the tools.
18. Terre’s favorite work tool is Granola AI for meeting notes. Being able to be fully present in a meeting, then prompt for specific details afterward, has changed how she absorbs information and follows up on decisions.
19. Terre is setting up her own Open Claude server at home to experiment with agents personally. She sees hands-on technical experimentation outside of work as essential for staying ahead of what’s possible and leading her team with credibility.
20. The progression of LLMs from their earliest versions to today has genuinely surprised Terre, even as someone who has worked in AI for decades. She points to the end-of-year memory summary feature as a small but striking example of just how contextually aware and useful these models have become.