Are You Learning the Right Skills for AI Product Management?
Are You Learning the Right Skills for AI Product Management?
Explore my TOP 7 challenges as an AIPM
#beyondAI
I fear that aspiring AI Product Managers are focusing too much on the AI aspect of the role. But is this really the case? If I were to grant you access to my inbox, you'd quickly discover that most of the questions I receive are about this.
However, for those looking to enter this field, merely possessing AI skills might not be sufficient. They might be surprised by the diversity of skills required. Since I cannot respond to everyone's questions individually, I intend to collect the most common ones and address them in a dedicated newsletter post. So, please forgive me if I don't reply immediately.
This issue is dedicated to helping you understand the challenges you will face as an AI Product Manager so that you can efficiently identify your deficiencies and build the necessary skills and experiences needed to thrive in this demanding yet rewarding field.
You will learn about:
A misleading perception: It’s all about AI
My 7 most frequent challenges in AI Product Management:
Problem
Justification
AI PM's role
Final Thoughts
Takeaway for AI PMs
Have fun 💡
A Misleading Perception: It’s All About AI
🛠 Expect more Engineering than Data Science problems in your AI Product Journey — Still, none reign as the most common challenge 🏆When discussing AI Product Management and the challenges we face, it might appear that we are primarily grappling with Data Science and AI-related issues.
This perception may be reinforced by 'AI' being the first thing you see in 'AIPM.'
Additionally, some of my (and others) LinkedIn posts may sometimes contribute to this kind of perception, even though I always emphasize that it’s #beyondAI 🤓.
But still, one might be tempted to think it’s all about AI.
Not a single word about all the other challenges - I’m sorry! At least I have this newsletter to make my points more comprehensible. 🤗Perhaps that's why many people interested in starting this new career believe that learning AI is the only step they need to take. This might be true for some who are really just one step away because they already worked in a similar field.
However, the questions I receive make me feel that I should at least share my thoughts with those asking just these kinds of questions:
What AI topics should I learn?
What AI skills do I at minimum need?
Can getting an AI certification help me get a job in AI Product Management?
There are so many AI topics; where should I begin?
Do I have to be an AI expert to start working as an AI Product Manager?
💡 I want to make it clear
The truth about AI Product Management tells a different story. A story that goes #beyondAI. And today, I will share my #beyondAI story with you. In this post, I won’t give you explicit answers to those questions above, but I assure you, by the end, you will definitely know what your next steps should be. 😉
My 7 Most Frequent Challenges in AI Product Management
Whenever I approach a new problem, I make it a point to understand it thoroughly. You might also believe that mastering AI is your only significant challenge. However, I want to provide a different perspective that will help you better define your questions and truly assess if understanding AI is indeed your sole and most pressing problem.
I will outline the seven most frequent challenges I encounter as an AI Product Manager. For each challenge, I will explain why it is common and discuss the role of an AI Product Manager in addressing it.
Below is the list of challenges you're likely to face throughout the AI Product Lifecycle, presented in ascending order:
7th Place
Data Science Problems (Initial Phase)
Problem Area: Data acquisition, cleaning, preprocessing, model selection, training, and validation.
In the beginning, it's really important to get the right data, as it lays the groundwork for creating effective models. Cleaning and organizing this data makes it useful and ensures it's good enough for building reliable models. Picking the right model, training it with your data, and thoroughly checking it are key steps that decide if the project will succeed.
Frequency Justification: These issues need a lot of attention at first but happen less often after the models start being used.
Setting everything up and getting the models ready takes a lot of work, but once the models are up and running, these tasks don't need to be done as often. This is because the big challenges of getting the models to perform well initially and making sure they work well with other systems are already dealt with.
AIPM Role: Leading the early stages of model development and making sure the models are strong and ready for use.
As an AI Product Manager, you're in charge of guiding the beginning phases, making sure everything meets technical and business standards. You'll work with data scientists and engineers to make sure the models not only work well but also fit the product's overall goals and meet customer needs. Your leadership is key in making sure everything goes smoothly as the project moves from building models to using them.
6th Place
Regulatory and Compliance Issues
Problem Area: Compliance with data privacy laws, ethical standards, and industry regulations.
Making sure that the product follows data privacy laws, stays ethical, and meets industry rules is very important. These are big areas that impact how safe and fair your product is.
Frequency Justification: These issues come up every now and then, usually when new laws are introduced or when significant updates to the product are made.
These challenges don't happen all the time, but they pop up when there are new rules to follow or when we make big changes and sometimes even small changes to our product and underlying model. It's like checking in to make sure everything is still in line with the law.
AIPM Role: Keeping the product legally safe, up to date with any changes in the rules, and handling legal complexities.
As an AI Product Manager, your job is to make sure that our product always meets the latest legal requirements and that we handle any legal issues wisely. You need to keep track of new laws and make sure the product changes to stay compliant.
5th Place
User Interface (impacting User Experience) Issues
Problem Area: Design usability, accessibility, and user feedback integration.
It's crucial to focus on making the design user-friendly and accessible to all, also when integrating new AI features into existing systems. This integration must be seamless and should consider how users interact with both old and new components of the system.
Frequency Justification: Although iterative, these issues are addressed during design sprints and post-launch feedback cycles.
We tackle these issues in cycles—planning, creating, testing, and improving the design based on continuous user feedback. This iterative process is particularly critical in environments where AI must work smoothly with existing technologies, ensuring that updates do not disrupt user experience or system stability.
AIPM Role: Partnering with UI/UX teams and system integration specialists.
As an AI Product Manager, your role involves not only working with UI/UX teams but also coordinating with system integration specialists to ensure that the introduction of AI into existing products is smooth and intuitive. This collaboration aims to maintain a balance between introducing advanced AI capabilities and preserving the usability and functionality of the existing system, based on direct feedback from users.
4th Place
Data Science Problems (Post-Deployment Phase)
Problem Area: Model monitoring, data drift management, model updating, and feedback loop integration.
Keeping an eye on how the AI models are performing, managing changes in the data over time, updating the models as needed, and using user feedback to make improvements.
Frequency Justification: These challenges keep coming up because the product continues to interact with users and process new data after it’s launched.
Once the product is out in the world, it doesn't stop evolving. We continuously face these issues because the product and its data environment are always changing, influenced by how users interact with it and the new data that comes in.
AIPM Role: Making sure AI models stay effective and updating them based on new insights and how users interact with the product.
As an AI Product Manager, your job is to ensure that the AI models continue to perform well over time. This involves adjusting the models in response to new data and user feedback to keep the product working well and meeting user needs.
3rd Place 🥉
Product Strategy and Market Fit Problems (also UX Problems)
Problem Area: Defining the product vision, aligning with market needs, feature prioritization, customer satisfaction, and enhancing user experience (UX).
It’s vital to clearly outline what the product aims to achieve, ensure it matches what the market desires, decide which features to develop first, keep customers satisfied, and continuously improve the user experience.
Frequency Justification: Ongoing adjustments are needed as markets and customer needs evolve over time.
As the marketplace and customer preferences change, it’s crucial to regularly update our product to remain relevant and engaging, including aspects of how users interact with it.
AIPM Role: Developing product strategy, analyzing market trends, and ensuring features not only meet user demands but also enhance user satisfaction through superior UX.
As an AI Product Manager, your job involves crafting a strategic direction for the product, staying ahead of market trends to anticipate future needs, and making sure that the product not only meets these needs but also delivers a satisfying and intuitive user experience.
2nd Place 🥈
Engineering and Data Engineering Problems
Problem Area: System integration, scalability, performance optimization, reliability, maintenance, security, and data engineering & infrastructure.
This area focuses on making sure all parts of the system work well together, can handle growing amounts of work, perform efficiently, stay reliable, are easy to maintain, stay secure, and have a strong data backbone.
Frequency Justification: These technical challenges are persistent, evolving with the product's lifecycle and technological advancements.
These issues are always present and change as the product grows and as new technology emerges. Keeping up with these changes is crucial for the product’s success.
AIPM Role: Working with engineering teams to resolve issues and maintain the product's technical health and scalability.
As an AI Product Manager, you are responsible for collaborating with engineering teams to tackle these technical challenges. Your role ensures the product remains technically sound, secure, and scalable as it evolves.
1st Place 🥇 Project Management Challenges
Problem Area: Resource allocation, timeline, and scope management, cross-functional team coordination, and risk management.
This covers managing how resources are used, keeping projects within set timeframes and scopes, coordinating teams from different areas, and handling potential risks.
Frequency Justification: Project management is a constant through all phases.
Managing projects is an ongoing need, essential at every stage - from the initial planning to the final maintenance of the project.
AIPM Role: Overseeing project progress, facilitating team collaboration, and ensuring projects align with business goals and (sometimes 😀) timelines.
As an AI Product Manager, your job is to keep an eye on how the project progresses, help teams work together smoothly, and make sure that every project meets the business objectives and stays on schedule.
Final Thoughts
Surprise, surprise 🎉 While the 'AI' in AI Product Management does give this role its unique flavor, at its core, it’s still very much about product management. That’s where the real challenges lie.
It's important to remember that these challenges can vary from one industry to another and from one type of AI product to another. For example, in the era of GenAI, we might see fewer interactions with traditional Data Science issues. This shift could alter the focus and skills needed in AI Product Management.
However, one common element remains: no matter the challenge, it usually involves some aspect of project management.
If you want to overcome a UI issue, your ability to effectively manage the project will be essential. Coordinating efforts between UI/UX designers, developers, and users to implement changes smoothly is key to success.
When facing an issue with model drift, project managing this issue will involve orchestrating a cross-functional team that includes data scientists, engineers, and business stakeholders to adjust and recalibrate the models accurately.
If there is a need for regulatory compliance or adaptation, managing this project will require close collaboration with legal and compliance teams to ensure all product updates meet the necessary standards without disrupting user experience or product timelines.
Project Management is often at the center of all the challenges you'll face.
This reflects my personal experience over the past 10 years as an AIPM, including nearly 5 years in a leading AIPM position at Vodafone, and I know it resonates with many AI Product Managers.
💡 But you are right!
AI knowledge is helpful - it actually helps a great deal. If you understand the technical underpinnings of how an AI product is built, you can steer its development more effectively. It allows you to make informed decisions, communicate better with your technical teams, and ultimately guide the product's direction to better align with business goals and user needs.So, you actually did ask the right questions. But the real question is, did you ask them wisely? I really hope so.And if not, you now know the importance of considering those beyondAI aspects 😉Takeaway for aspiring AIPMs
I hope this list underscores the essence of working as an AIPM and highlights the competencies you should hone to make a significant impact.
Now you have my personal list of the 7 most frequent challenges. Use this list to identify what you need to learn next.
Here’s how to do it in 7 steps:
Review Each Challenge: Start by carefully reviewing each challenge listed. Understand the specifics of each and how they relate to your current skills and knowledge.
Self-Assessment: For each challenge, assess your current level of expertise. Are you a beginner, intermediate, or expert in that area? Be honest with yourself about where you stand.
Identify Gaps: Compare your current skills against the requirements of each challenge. Note the areas where your skills are lacking or where deeper knowledge could help you perform better.
Set Learning Goals: Based on the gaps you’ve identified, set specific, measurable learning goals for yourself. Decide what you need to learn and set a timeline for achieving these goals.
Seek Resources: Look for resources that can help you achieve your learning goals. This could include books, online courses, workshops, or even finding a mentor in the field. Some people have even mentioned that my newsletter and LinkedIn posts are helpful—don't believe them, though! 😂
Apply What You Learn: As you acquire new skills, find ways to apply them in your current role or projects. Practical application will reinforce your learning and demonstrate your growing capability.
Reassess Regularly: Learning is a continuous process. Make it a habit to regularly reassess your skills and update your learning plan as you grow and as new technologies and methodologies emerge in the field of AI Product Management. This advice isn't just for those starting out; it applies to everyone in the field, myself included! 🤓
Stay curious.
Stay driven.
And you'll find your path to becoming an effective AI Product Manager.
JBK 🕊
P.S. If you’ve found my posts valuable, consider supporting my work. While I’m not accepting payments right now, you can help by sharing, liking, and commenting here or on my LinkedIn posts. This helps me reach more people on this journey, and your feedback is invaluable for improving the content. Thank you for being part of this community ❤️.
Your Feedback Matters
Whether you're a seasoned AI Product Manager or just starting out, how did this list resonate with you?
Aspiring AIPMs: Is this list helpful in guiding what you need to tackle next?
Experienced AIPMs: Are there other challenges you encounter more frequently?