Product Management Is an AI Candy Store
FOR IMMEDIATE RELEASE
Product Management Is an AI Candy Store
Product management is the best function to start your AI transformation.
According to a 2023 study by Accenture, 75% of C-suite executives agree that failing to effectively integrate AI within the next five years could result in business obsolescence. I predict that the 25% of leaders who don’t believe in that future will likely be the first to become obsolete.
AI is poised not just to support decision-making or process data but to lead in driving product innovation and strategic planning. The field of product management is particularly well-suited for AI applications due to its data-rich environment and dynamic nature. AI’s ability to analyze large data sets, identify patterns, predict outcomes, and automate tasks aligns well with the needs of modern product management and strategy.
Here’s why product strategy, management, marketing, and supporting functions are prime candidates for AI integration:
Structured, Repetitive Tasks
Product management and strategy involve numerous repetitive and time-consuming tasks—such as market research, competitor analysis, and design asset development—that AI can automate. This automation enhances efficiency and allows human teams to focus on more strategic and creative aspects of their roles.
Complex Decision-Making
The functions involve complex decision-making processes that require the analysis of vast amounts of data, including market trends, consumer behavior, competitive analysis, and financial projections, for example. AI excels at processing and analyzing large datasets quickly and accurately, offering insights to inform strategic decision-making.
Predictive Modeling
Predicting market trends, customer preferences, and product performance is invaluable in our field. AI can forecast future outcomes based on historical data, enabling product managers and strategists to make proactive decisions.
Probabilistic Outputs
The nature of what practitioners produce in these functions tends to be more probabilistic than deterministic. From market sizing to effort estimation to financial projections, all tend to be analyzed in ranges and probabilities. This makes the field a perfect playing field for today’s AI model outputs, which tend to be more probabilistic.
Continuous and Fast Change
Markets are dynamic, with customer preferences, technology, regulatory, and competitive landscapes constantly evolving. Implemented correctly, AI systems can continuously monitor these changes and help you adapt strategies in real time. Your solutions cheat sheets or competitors’ battle cards that used to be updated semi-annually or quarterly need to be updated in close to real time, and AI can help.
Macro and Micro
In addition to performing all this analysis at a macro, industry, and competitor level, the same needs to be done for individual customers, customer persona, or segment. Understanding and addressing individual customer needs are crucial for the success of any product. AI can analyze customer data at scale to uncover insights into customer preferences and behaviors, allowing for the creation of personalized experiences and products that better meet market demands.
Less Reliant on Sensitive People Data
Fields like HR and finance utilize large amounts of sensitive and private people data, making training and implementing AI models more complex. Although we utilize a lot of confidential and trade-secret data, implementing AI models in our field is less sensitive. At a minimum, it is unlikely to run afoul of regulations like GDPR or the New York AI Hiring Law, which comes with steep consequences.
Large Volumes of Data
The more training data available, the happier data scientists are. Product-related data is everywhere. The ETL may be more complex than it is with more structured forms of data like finance or supply chain data, but it is feasible.
Last but not least, and it may be the top reason AI will thrive in product strategy and lifecycle management: YOU! If you are reading this, you are likely a forward-looking, education-hungry, innovation-seeking entrepreneur or product leader. You are interested in learning and are open to change. You likely invoke, seek, and thrive on change while others avoid it. This describes the majority of the practitioners in our field. The same can’t be said of many other fields for good reasons related to the specific circumstances of these segments. Because of the nature of the people involved in this field, AI will be adopted early and will thrive here.
Although the field is an excellent match for AI, that doesn’t necessarily mean that adopting AI will be easy. We will likely be disappointed if we approach AI adoption like any traditional technology adoption or transformation project. Some of the hurdles that we need to prepare for include:
Complex and Time-Consuming
Whether you are training and building your own models, leveraging commercially available foundation models, or implementing a new product that leverages AI, the process is complex and time-consuming. To succeed in any of these scenarios, technical aspects (data and compute capacity) and organizational change need to be addressed. Forecasts on time, budget, and TTV need to be realistic.
Talent Availability and Bandwidth
Much is being discussed about the future of work and the KSAs (Knowledge, Skills, Abilities) we will need. Whether acquiring this talent, outsourcing the project, developing from within, or a mix, you need to realistically account for how long it will take you to form your AI team.
Moving Target
This is likely the biggest concern. Given how fast this technology is advancing and changing, there may be better options than what you assess, select, and implement today. Organizations will have to build AI roadmaps that account for future capabilities that we may not even imagine today. More on this topic below.
Because of these hurdles and the other general ones around AI (questions of ethics, data security, privacy, bias, future of humanity, job security, etc.), taking full advantage of AI will likely be slower than optimal for many organizations. Many organizations will be stalled at implementing AI as a tool vs AI as a co-pilot or agent, which is the topic of the next section.
