CPO Rising Series: Comcast Product SVP on Rewiring the Entire Product Lifecycle for AI

What happens when software powers every customer-facing interaction a business has, from the app to the agent interface to the care experience? Everything becomes a product. In this episode of the CPO Rising series, hosted by Products That Count Resident CPO Renee Niemi, Comcast SVP of Connected Living Products Randall Hounsell speaks on what it means to lead product at one of the most complex consumer technology ecosystems in the world. Randall shares why data-driven is misunderstood, how he is rewiring Comcast’s entire product lifecycle to be AI-first, and why vision and strategy are the most durable skills a product leader has in a world where everything else is changing fast.

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. Everything is a product. Software powers every customer-facing interaction a business has, from consumer apps to agent interfaces to care experiences. The sales team is a product. The support team is a product. The process of building, launching, and optimizing all of it is the same craft.
2. Data-driven has been misrepresented. If every decision you make is based on an A/B test or what the data already confirms, you are an optimization machine, not a visionary. Data is how you navigate toward a vision. It cannot replace having one.
3. Vision-led leadership is what separates great product leaders from good analysts. At the senior level, the biggest decisions require massive extrapolations from incomplete data. The vision is what makes those extrapolations coherent. Without it, data becomes noise.
4. The entire product lifecycle needs to be rewired as an AI-first lifecycle. This is a multi-year journey, not a feature toggle. From gathering analytics to prototyping to writing requirements to coding to testing to post-launch optimization, AI should be doing the lifting while product leaders stay in the decision seat.
5. Agile reduced documentation in favor of communication. AI reverses that. If AI is doing the coding, you need zero ambiguity in what you are building. Fully speccing and documenting the product is now more important than ever, not less. That is a significant shift in how teams need to operate.
6. Platforms are becoming thicker, and that is a competitive advantage. Cloud and mobile forced the first wave of platform leverage. AI and LLMs are adding another layer. Companies that build platforms where the incremental cost of adding a new product or feature approaches zero are the ones that will move fastest.
7. The product manager to engineer ratio could shift from 1:10 to 1:1. As AI takes on more of the coding, testing, analytics, and design work, a single product manager can own more of the full stack. Jobs do not disappear. The skill sets required to do them change fundamentally.
8. Agents do not correspond to jobs. Jobs are based on what a person can do. An agent can do a function without being a person. Once you stop mapping agents to job titles and start mapping them to functions, the entire operating model becomes more flexible, more powerful, and less constrained by headcount.
9. Vision and strategy are the most durable skills in the AI era. AI is good at observing and gathering facts about the world that exists. It is not good at envisioning a world that does not exist yet. Articulating the vision and getting everyone behind it remains the core job of a great product leader, and that has not changed.
10. Teams need to own the full lifecycle to be truly accountable. Understanding the market, owning discovery, driving implementation, and measuring outcome are not separate jobs. They are one job. Teams that are empowered and skilled to run that full lifecycle are the ones that can be genuinely accountable for the vision.
11. Innovation within the current business model is different from innovation of the business model itself. The former can happen within existing teams if they are structured correctly. The latter may require separation. Knowing which one you are doing is the first decision, and most companies confuse the two.
12. Real-time data pipelines are one of the highest-leverage investments a product leader can make. Knowing what customers are doing, where they are struggling, and what they need next, in real time, is what makes personalization possible. Without the data infrastructure, the AI has nothing to work with.
13. The goal is an ecosystem of products that are better together. The most powerful moat a connected product company can build is not any single feature. It is the compounding value of products that perform better in combination than they do alone. Comcast’s mobile and broadband bundle was built on exactly that hypothesis.
14. Build versus buy is more nuanced than it appears. Acquisition looks easier from a distance than it is up close. Integrating technology built for a different world in a different context is genuinely hard. Many times, building exactly what you want and being patient to get there is the right answer.
15. Each generation of technology gets faster and more impactful. The internet gave us Meta and Nvidia. AI is the next exponential shift. The companies and product leaders who structure their teams to live in that accelerating world rather than react to it are the ones who will define the next wave.
16. What makes your product unique is what you should be most focused on. As platforms and LLMs commoditize more and more of the infrastructure, the imperative is to isolate the things that only you can do and invest everything there. Leverage the common. Own the unique.
17. AI in the network means self-healing infrastructure, not just better apps. The vision at Comcast is a network that identifies issues automatically and resolves them before the customer notices. That is a fundamentally different relationship between infrastructure and experience than anything that has existed before.
18. Device identification is the foundation of personalized connectivity. Before you can optimize Wi-Fi, manage screen time, or prioritize devices, you have to know what is on the network. AI solving that identification problem at scale is what makes the entire personalized experience possible.
19. Temporary job reduction may happen. Long-term displacement is not the right frame. In the short term, some roles will contract as AI takes on more functions. Over time, the skill sets will evolve and the roles will transform. The leaders who help their teams make that transition rather than fear it will build stronger organizations.
20. The human’s most irreplaceable contribution is envisioning what does not yet exist. AI synthesizes everything that is already known. The net new, the zero to one, the genuine creative leap, that is still entirely human. And as AI takes on more of the execution, the percentage of time humans spend on that creative work will grow dramatically.