Evolving PM: Navigating Seismic Shifts in Game Dev – Insights from Product Leaders
Major macro changes have disrupted and continue to disrupt the gaming industry, as PMs we are experiencing seismic changes to our industry, organizations, teams, and the PM role
The game development industry is in a period of unprecedented flux, a whirlwind of fierce competition, rapidly evolving player expectations, and the groundbreaking, almost daily, advancements in technologies like Artificial Intelligence. For studio executives, product managers, and marketers navigating this landscape, understanding these shifts isn't just beneficial—it's critical for survival and success.
During GDC, AWS and CleverTap sponsored a Product x LiveOps Symposium event, including a panel, "The Evolution of the Product Manager," dissecting the significant industry changes and their implications for the role of PM.
This post distills some of the most compelling observations and actionable takeaways from these discussions. It will explore the macro forces reshaping game development, the tangible impact of AI (both current and future), the shifting identity and skillset required of product managers, and crucial strategies for thriving in this dynamic new era.
Watch the full video below, including an after-panel discussion with me, AWS (Tim Hong) + CleverTap (Solomon Lichter) below! 👇👇👇
🎧 Listen on Spotify, Apple Podcasts, or Anchor
Speakers:
Oren Debi. Generative AI Lead & VP Product at SciPlay.
David Duong. Sr. Director of Product Management at Respawn Entertainment (Apex Legends).
Tim Hong. Head of Live Service Games at AWS for Games.
Lee Horn. Fmr. VP of Product & Game Director at Mountaintop Studios.
Joseph Kim. CEO at Lila Games.
Solomon Lichter. Sr. Director, Global Gaming at CleverTap.
Part 1: Decoding the Product x LiveOps Panel – The Product Manager in Flux
We are faced with an industry undergoing significant change, which is forcing the evolution of the product management role within it.
The New Battlefield: Macro Trends Redefining Game Development
Several powerful macro trends are fundamentally altering the terrain for game developers and product managers (PMs). These aren't isolated phenomena but rather interconnected forces that create a complex new competitive environment.
The Attention Economy Squeeze: A primary challenge is the "attention economy." In the discussion, we highlighted that "getting attention, especially when launching a new game, has become increasingly difficult" due to the sheer volume of competing products vying for a user’s time, including video-on-demand, social media, TikTok, and the growing dominance of popular, live-operated games. This isn't merely about an increase in the number of games; it signifies a fundamental shift in how players allocate their finite leisure time. For executives and PMs, this translates into an arduous uphill battle for visibility and initial player traction for new titles. Consequently, more innovative and potent marketing and engagement strategies are demanded from the very conception of a game. The scarcity of attention directly influences return on investment calculations for new intellectual property. Studios may need to anticipate higher user acquisition costs or accept longer payback periods, which in turn can sway decisions about which projects get the green light, potentially favoring more established IP or less experimental concepts.
The Colossus of Global Competition: The "rise of China" continues to impact our industry as a formidable competitive force, with "extremely big-budget, huge teams working 996," making it "very difficult to compete against them." David Duong from EA, working on Apex Legends, reinforced this by citing the example of Marvel Rivals, which is launching four new characters in a single season—a content velocity he described as "insane" and something that has "massively raised player expectations." This isn't just about more competitors; it establishes a new global benchmark for content volume and production cadence. Western studios now find themselves contending with development entities operating under vastly different economic and labor paradigms, and this directly reshapes player expectations worldwide. This intense competitive pressure, particularly concerning the speed of content delivery, may serve as a catalyst, accelerating the adoption of AI tools for content creation and workflow optimization not merely as an enhancement but as a fundamental competitive necessity. If competitors can produce vast amounts of content rapidly, maintaining player engagement without similar output becomes a significant challenge, pushing studios towards AI solutions to bridge the gap.
LiveOps: The Resurgent King: A "resurgence in LiveOps" was one of the biggest trends of last year, when games like Brawl Stars reportedly increased their revenue sixfold: a significant revelation of what’s possible for even older live-operated games. This success underscores a fundamental question in today's market: "How do you succeed in new game development when it's so challenging?" The sustained triumph of long-tenured LiveOps games provides a validated model for achieving enduring revenue streams and player engagement. For PMs and studio executives, this observation reinforces the critical importance of long-term product strategy, continuous and meaningful content delivery, and robust community management, moving away from a "launch and leave" mentality. The pronounced difficulty in successfully launching new games, stemming from the crowded attention economy and fierce competition, makes investing in and nurturing existing successful LiveOps titles an increasingly attractive, and potentially less risky, proposition for studios. This can lead to a strategic reallocation of resources within organizations, prioritizing the known quantities of successful live games over the uncertainties of new ventures.
Navigating Incumbents and Escalating Player Expectations: Lee Horn offered a stark warning: "Do not go up against an incumbent that has been around a long time," especially in competitive PvP genres that boast "years of credibility with high-quality matchmaking and great anti-cheat." He emphasized that in the current market, "a small innovation is not enough." David Duong added that established titles like Fortnite and Call of Duty present an "onslaught of content," compelling other games to "meet players where they need us to be so they don't go to other games." The barrier to entry and, more importantly, to sustained success in established genres, particularly player-versus-player experiences, is now extraordinarily high. Players have come to expect a baseline of polish, content depth, and technical stability that new entrants must not only meet but often exceed almost immediately upon launch. This market dynamic tends to favor either highly differentiated new IPs that carve out unique niches, games that can generate entirely new market segments, or titles that benefit from significant pre-existing IP leverage. Incremental improvements within already crowded spaces are proving increasingly insufficient to gain traction, demanding that PMs conduct brutally honest and rigorous market assessments before committing to new projects.
AI in the Trenches: Separating Hype from Reality
Artificial Intelligence was a central theme, with panelists keen to move beyond speculative discussions to its real-world impact on game development.
AI: Transformative, Not Transitory: Tim Hong of AWS asserted that AI "is now a significant development, not a fad." Oren Debi from SciPlay concurred, noting that while truly transformative, game-changing uses of AI are not yet fully realized, its "adoption is mainly around efficiency" in current workflows. This consensus establishes AI as a fundamental technological shift that demands strategic attention from industry leaders, signaling that investment in AI capabilities is evolving from an optional experiment to a strategic imperative.
Current Practical Wins with AI: The panel highlighted several areas where AI is already delivering tangible efficiency gains:
Art Creation: Oren Debi reported "about a 40% increase using AI tools to create collectibles and other features."
Code Assistance: "Code Assistant is also greatly used," according to Debi.
PM Productivity: Large Language Models (LLMs) are being employed by PMs for "competitive intelligence, PRDs, and wireframes—all the heavy stuff that takes a long time."
QA Testing: Although discussed more in the after-panel, its relevance to current utility was clear, with AI advancing QA "beyond simple automation to contextual automation."
These concrete examples demonstrate that AI is not just a future promise but a present-day tool capable of delivering immediate ROI in specific departments. For PMs, this means an opportunity to explore these tools to automate or accelerate tasks, thereby freeing up valuable time for more strategic, high-impact work. The pattern emerging from these current applications is that AI is largely augmenting human capabilities and streamlining repetitive or time-intensive tasks, rather than outright replacing roles.
The "Amplify, Not Replace" Consensus (for now): A key sentiment from the panel was that AI's current role is one of augmentation. David Duong stated, "I think AI is here to amplify, not replace." Oren Debi delivered a more pointed message: "The only people who will be replaced are those who don't adopt AI. PMs who adopt AI will eventually replace PMs who don't." This offers both reassurance and caution: AI tools are rapidly becoming essential, and proficiency in leveraging them will be a significant differentiator. The immediate concern isn't AI taking jobs but rather PMs who effectively integrate AI into their workflows, outperforming and potentially displacing those who do not. However, this "amplify, not replace" stage might be transitional. As AI capabilities continue to advance rapidly, particularly towards more sophisticated forms of narrow AI or even nascent Artificial General Intelligence, the distinction between amplification and replacement may become increasingly blurred. This is especially true for tasks currently considered core components of certain roles.
The Evolving Product Manager: Bridging Worlds and Skills
The discussion extensively covered how the product manager role itself is transforming in response to these industry and technological pressures.
The Great Divide: Mobile Agility vs. PC/Console Tradition: Solomon Lichter, self-described as a "mobile guy," highlighted that mobile PMs are "very sought after due to their skill set, their ability to act on data, and their general working style." He traced mobile's DNA back to tech giants like Facebook and Zynga, fostering a culture of "move fast and break things." Conversely, Lee Horn pointed out that the PC and console space has a "long history of succeeding without product managers or data insights," which has cultivated a degree of "internal resistance" to their integration. This illuminates a significant cultural friction point within the broader games industry. As games across all platforms increasingly adopt LiveOps models and data-driven decision-making, the agile, iterative mindset characteristic of mobile PMs becomes more valuable. Yet, traditional AAA organizational structures often struggle to integrate these PMs and their methodologies effectively. This resistance within PC and console environments to data-informed PM practices can result in slower adaptation to market changes, missed optimization opportunities, and overall inefficiency, particularly as market pressures intensify the demand for greater agility. The "education process" Lee Horn mentioned is therefore not just a minor adjustment but a critical undertaking for the future competitiveness of many AAA studios.
Hiring for the AAA Future: Fundamentals, Empathy, Communication: When discussing hiring PMs for the AAA space, David Duong emphasized the importance of "sound fundamentals," asserting that if a PM possesses a strong foundational skillset, "they can solve and do anything." Beyond core competencies, he stressed the critical need for "strong empathy, especially with designers," who, as he noted, "don't always respond well to data." This necessitates an ability to "read the room, understand what they care about, and think on your feet to reposition your message to connect with them and get things done." In the complex, multi-stakeholder environments typical of AAA development, raw technical or analytical skills are insufficient. PMs must possess robust foundational problem-solving capabilities, coupled with high emotional intelligence, to effectively navigate internal team dynamics and drive alignment. This is particularly crucial when introducing data-driven approaches into cultures traditionally led by design intuition. The emphasis on "empathy" and the skill of "repositioning messages" suggests that a core competency for PMs in the AAA sector is internal diplomacy and effective change management, especially when data-driven findings challenge established "tribal knowledge" or long-held artistic intuition, a point Solomon Lichter also touched upon when describing console organizations.
AI's Shadow: Team Compression and the Rise of the "PM+1" As an "AI maximalist," (sort of), I offered a somewhat contrarian perspective on the impact of AI. I discussed how AI will "fundamentally subsume a lot of work from different roles," which will inevitably lead to a compression of traditional roles and the emergence of hybrid positions such as a "product engineer (PM + engineer) or a product marketer (PM + UA marketer)." To encapsulate the necessary evolution for product managers in this AI-suffused future, I sated how I believe the role of PM should be "PM+1": a product manager who not only masters the core PM skillset but also possesses "extreme depth in another area," be it engineering, marketing, design, or data science. This vision suggests that generalist PMs, whose value is tied to tasks that AI can increasingly automate, may find their distinctiveness diminishing. Specialization or profound expertise in a complementary field will likely become crucial for differentiation and sustained value. This outlook directly challenges conventional organizational structures built upon clearly demarcated, siloed roles. Companies that can successfully cultivate, hire, or retrain for these hybrid "PM+1" talents, and that can foster flexible "organizational metas" will be better positioned for adaptability and competitive advantage in the evolving landscape. If AI handles many current PM tasks, humans must bring skills that AI either cannot replicate or can uniquely leverage, leading to this more specialized, deeply skilled PM profile.
Key Panelist Takeaways: Your Evolution of PM Cheat Sheet
The panelists each offered a concise, potent piece of advice for navigating the changes ahead. These takeaways provide a clear roadmap for product managers and executives alike.
Part 2: The After-Panel Huddle – Deeper Reflections & Future-Proofing
Recently, Tim Hong, Solomon Lichter, and I reconvened for a debrief, delving deeper into some of the themes and sharing further reflections on the rapidly evolving landscape.
Beyond the Stage: Unpacking Surprises and Lingering Questions
The follow-up conversation revealed a palpable sense of both excitement and uncertainty pervading the industry.
Industry's Thirst for Tactical AI Knowledge: Solomon Lichter observed a "huge hunger for more insights—tactical, real-world, and strategic" concerning AI, evidenced by "many follow-up requests to drill into specific areas or set up calls" after the panel. Tim Hong expressed that he was "positively surprised by how folks are thinking about AI and its possibilities," noting a clear progression in the industry "from idea to true POCs [Proofs of Concept], and possibly even into production this year." This indicates that the game development sector is moving beyond general discussions about AI's potential and is now actively seeking practical implementation strategies, best practices, and collaborative avenues to harness its power. The expressed interest in "collaboration" and clients wanting to build "councils" suggests a recognition among studios of the inherent complexity of AI implementation. Rather than attempting to develop all necessary AI capabilities in-house, many are looking towards partnerships and shared learning ecosystems, potentially opening significant opportunities for specialized AI service providers and consultants.
Embracing Uncertainty in a Sea of Change: One observation I noted is about "the level of uncertainty and the embrace of change I'm seeing among leaders in the gaming industry." Many executives I talk to are "unsure what these environmental changes mean for them and how to prepare for a world with so many concurrent developments," especially with the significant looming impact of AI. From what I’m hearing, even seasoned industry leaders are grappling with the sheer pace and confluence of transformative forces—the lingering effects of the COVID boom, shifts in the attention economy, and the pervasive rise of AI. In such an environment, characterized by high uncertainty, the value of robust frameworks for decision-making, opportunities for shared learning (such as industry panels and discussions), and the adoption of agile, adaptive processes increases dramatically. Static, long-term strategic plans become inherently less reliable, necessitating a more flexible and responsive approach to leadership and planning.
AI: From Feature to Foundational Orchestration
The after-panel discussion further refined the understanding of AI's evolving role, positioning it not merely as a set of features but as a fundamental orchestration layer.
The PM's Evolving Value Proposition with AI: Tim Hong posited that as basic knowledge acquisition becomes "somewhat commoditized by AI"—for instance, quickly getting context on an unfamiliar term—product managers will "need something else to hang your hat on, whether it's being design-focused or analytics-focused." The critical question becomes, "What will get you into the room to speak to issues and offer opinions?" If AI tools can efficiently handle foundational research and data queries, the onus falls on PMs to elevate their contributions toward more strategic thinking, specialized expertise, and the uniquely human elements of judgment, creativity, and experience informed by data. This directly implies that PMs whose primary value lies in tasks easily automated by AI must actively develop deeper, less automatable skills to maintain their relevance and impact.
Offensive & Defensive AI: Automation Meets New Capabilities: PMs should consider how AI can automate existing work (a defensive posture, focused on efficiency) and, crucially, how it enables entirely new capabilities such as advanced personalization, dynamic offers, and sophisticated CRM (an offensive posture, focused on value creation). This offensive application of AI can empower "smaller teams... to compete with larger ones" by leveraging these new, powerful capabilities. This framing is vital because it positions AI not just as a means to cut costs or accelerate existing processes, but as a catalyst for unlocking entirely new avenues for player engagement and value generation, potentially re-leveling the competitive playing field.
The PM as Orchestrator and Tastemaker: In the world we are moving towards with AI, a significant mindset shift is required towards leveraging AI less as a discrete feature layer and more as an "orchestration layer." In this paradigm, the PM is envisioned "at the controls of a command center orchestrating everything." Significant PM value may come as a "tastemaker for capabilities and tools," responsible for deciding "which specific actions to take to optimize the game" from the myriad possibilities AI enables. This perspective elevates the product manager's role from that of a feature owner or backlog manager to a strategic integrator and optimizer of complex, AI-driven systems. Such a role demands a broad understanding of AI's potential across various domains, coupled with the acuity to direct its application in the most effective and impactful manner. If PMs are to become orchestrators of diverse AI tools and agents, their core competencies must evolve to include strong systems thinking, the ability to strategically prioritize AI initiatives based on potential impact, and a keen understanding of the ethical considerations and player experience implications of AI-driven decisions. A high-level technical understanding of AI also becomes increasingly important in this context.
Gartner's "3 Returns" Framework: Strategic AI Investment: Tim Hong introduced a valuable framework from Mary Maseglio, a VP at Gartner, designed to help leaders understand and categorize AI business outcomes. The "Three AI Business Outcomes Every Leader Must Understand" are: "return on employee" (focusing on enhancing employee efficiency and productivity), "return on investment" (aiming to directly improve the bottom line or key performance indicators), and "return on future" (involving the creation of entirely new capabilities and business models). This framework provides a structured approach for executives and product managers to categorize, justify, and measure their AI initiatives. It allows organizations to move beyond vague notions of "investing in AI" towards specific, outcome-oriented projects with clear objectives and metrics. For instance, Tim Hong described a project to leverage AI to create a comprehensive knowledge base capturing all past initiatives a studio has tried on a game. Such a project clearly falls under "return on employee," with its value measured in efficiency gains, reduced redundancy, and accelerated onboarding, rather than direct, immediate revenue impact.
"Product Velocity" as a North Star Metric: A key metric that will become increasingly important is "product velocity." This may become "the most important metric for all PMs," particularly within the mobile and live ops spheres. In a market characterized by rapid evolution and intense competition, the speed at which teams can iterate, learn from those iterations, and deploy valuable changes to players becomes a critical competitive differentiator. Artificial Intelligence is seen as a major enabler of this necessary increase in velocity. Further, AI turns product velocity into a bigger point of differentiation between those who leverage AI for velocity and those who don’t. Further, the focus on product velocity directly addresses the competitive pressures previously discussed, such as the high content output from global competitors and the imperative to meet rapidly changing player expectations. If the market and competitors are moving fast, and AI can accelerate internal processes, then measuring and optimizing for product velocity becomes a crucial mechanism for effectively leveraging AI and maintaining a competitive edge.
Agentic AI: The Next Frontier in Live Ops and Beyond
The conversation then ventured into the more advanced concept of agentic AI, painting a picture of a future where AI plays an even more autonomous and proactive role in game operations.
Understanding Agentic AI: AI That Learns and Evolves: Tim Hong provided a simple yet effective explanation of agentic AI: "If you let this 'person' keep playing the game, after a while, they're going to get better at it." This distinguishes agentic AI from more static AI models, implying systems that can autonomously learn from their environment and interactions, adapt their strategies, and optimize their performance over time. Solomon Lichter underscored the potential impact of this technology, viewing agents as an "existential shift" that will eventually "completely consume automation" as we currently understand it, particularly in marketing automation and, by extension, LiveOps. This represents a significant step beyond current AI applications, envisioning AI systems that operate with a greater degree of autonomy within defined parameters, which has profound implications for how games are managed, personalized, and monetized.
Practical Applications: A Glimpse of Functional Agents: The discussion explored several potential applications of these functional, learning AI agents within game development and live operations:
Segmentation Agent: An agent "whose only job is to look at how to segment players," continuously refining segmentation based on evolving data. (Tim Hong)
Non-Spender Health Agent: A particularly insightful concept was an agent dedicated not to aggressive conversion, but to "observe the health of non-spenders... to figure out if we're engaging our player base in the most positive, healthy way." (Tim Hong) This highlights a shift towards using AI for holistic player experience management.
In-Game Economy Management: Agents could be tasked with "actively managing the in-game economy in real time," responding to shifts in currency availability, player pools, and spending habits to maintain balance and health. (Solomon Lichter)
Offer Tuning & Personalization: Agents could dynamically tune offers, enabling "true one-to-one personalization" at a scale unachievable through manual methods. (Solomon Lichter)
Dynamic Content/Asset Tagging: AI could dynamically tag in-game content and assets, a typically manual and laborious process, leading to significant efficiency gains. (Solomon Lichter)
These examples illustrate how specialized AI agents could potentially take over complex, dynamic LiveOps tasks that currently demand considerable human oversight and intervention. This could lead to more deeply personalized, responsive, and optimized player experiences delivered at an unprecedented scale. The emergence of such functional agents would necessitate a shift in the product manager's focus: from the direct execution of these LiveOps tasks to defining the goals, constraints, and ethical guidelines for these AI agents, and then diligently monitoring their performance and overall impact. The PM effectively becomes a "manager of AI agents."
The Unshakeable Foundation: A Rock-Solid Data Strategy: A critical prerequisite for realizing the potential of advanced AI, especially agentic systems that learn and adapt based on data, is an impeccable data strategy. Tim Hong stated unequivocally, "If your data strategy is not solid, it's really hard to create a training model on that." This cannot be overstated. The efficacy of these sophisticated AI systems is entirely dependent on the quality, comprehensiveness, and accessibility of the data they are trained on and operate with. "Getting your data house in order," as Tim put it, is a non-negotiable foundational step before embarking on ambitious agentic AI initiatives. A deficient data strategy will inevitably lead to poorly trained AI models, resulting in ineffective or even detrimental AI agents, ultimately undermining the entire AI initiative.
The Culture Clash Revisited: Adaptability as the Ultimate Competitive Edge
The after-panel discussion returned to the persistent challenge of cultural integration and the overarching importance of organizational adaptability.
The Persistent Struggle: Integrating Mobile DNA into Traditional Studios: Solomon Lichter reaffirmed the "cultural and organizational dichotomy" between the agile, data-centric world of mobile game development and the more traditional structures of many PC/console studios. He observed that while console companies actively want to hire mobile PM talent, recognizing the value of their skills, these organizations are often "not set up organizationally, culturally, or politically to do so." This frequently leads to a "retention issue," as mobile-native PMs become frustrated by slower paces and resistance to rapid, data-informed iteration. This is not merely a recruitment challenge; it points to a need for fundamental organizational change. Studios that cannot evolve their internal cultures to genuinely embrace and empower agile, data-driven, and velocity-focused methodologies will find it increasingly difficult to attract, retain, and effectively leverage top product talent, ultimately hindering their ability to keep pace with market demands.
Failure to Adapt = Losing the Velocity War: "If you dig in, remain narrowly focused, and stick to 'this is how we did it before,' you'll struggle." Solomon Lichter connected this directly to the AI discussion, noting that if organizations are already finding it challenging to implement current LiveOps best practices, it will be "even harder for them to adapt to a new reality where human intervention... will be increasingly mitigated" by AI. The accelerating pace of change, significantly amplified by AI's disruptive potential, means that cultural inertia and resistance to new ways of working are greater liabilities than ever before. The ability to adapt quickly—encapsulated in the term "product velocity"—is rapidly becoming a primary determinant of competitive success. The "adapt or die" message is not mere hyperbole in this context. The potent combination of ongoing market shifts and AI's transformative capabilities will inevitably create clearer distinctions between winners and losers, largely based on their respective organizational adaptability and willingness to evolve. Traditional studios that fail to address these cultural and organizational impediments will find themselves increasingly outmaneuvered in an AI-driven future that prizes speed and data-driven agility.
Conclusion: Your Roadmap for the Future of Game Product Management
All of these discussions highlight the game development landscape is undergoing a seismic transformation. While the role of the product manager is arguably more critical than ever in navigating this complexity, its fundamental nature is changing. It's becoming less about the execution of specific, narrowly defined tasks and more about strategic orchestration, the cultivation of deep specialized knowledge, and the championing of organizational adaptability.
A final practical trinity of advice for thriving in this new era:
Get Hands-On & Be Curious: Tim Hong urged PMs to actively engage with new technologies, emphasizing that "The barrier to entry... is extremely low... Go take a ride, take a spin, experiment." In a world where tools and platforms are rapidly evolving, particularly in the AI space, practical, hands-on experimentation is invaluable for building intuition, understanding capabilities and limitations, and identifying opportunities.
Build in Public & Learn Out Loud: Solomon Lichter advocated for a shift away from secretive development cultures, advising PMs to "Share your experiments, failures, and thinking openly," because "The future likely belongs to PMs who learn out loud." In the face of complex, rapidly advancing technologies like AI, collaboration and shared learning will accelerate both individual growth and industry-wide progress. Openness can foster a collective intelligence that benefits all.
Embrace Continuous, Urgent Learning: "The impact of AI will be the most significant not only in our lifetime but possibly in all of human history. Now is not the time to relax... learn as much as possible and push yourselves now." In these coming years, there may be a "power-law distribution of value" and wealth. Hence, in this new landscape, being slightly ahead of the curve could yield disproportionately large rewards.
The future of game product management is undoubtedly challenging, demanding new skills, new mindsets, and a new level of adaptability. However, it is also a future brimming with opportunity for those who are willing to embrace change, commit to continuous learning, and lead with strategic foresight. Hopefully, our discussions offer a valuable compass for navigating the exciting and transformative journey ahead.
Good luck!