“AI” is NOT going to fix Product Management. But you can. Here’s how.
Will AI replace Product Management? No. But it will impact our work. And to leverage it well, we have to fix Product Management first. Are you up for it?
Contents
- The Potential Impact of “AI” on Product Management
- What Problem are we Solving
- What are the Problems we Should be Solving?
- Does “AI” help with any of this?
- Product Leaders Must Step Up
- How to (re)DESIGN your Org
- A Little Feedback Please
The Potential Impact of “AI” on Product Management
There’s a lot of talk — some of it well justified — about the impact that “AI” will have on knowledge workers.
NOTE: I put “AI” in quotes, because, the term “artificial intelligence” was coined in 1956, almost 70 years ago, by John McCarthy. It was an aspirational description for technology that would, one day, allow machines to “think” and “reason” in ways or with capabiltiies analogous to what humans do.
Clearly we have not yet reached that state today, so I use “AI” in quotes to clearly designate that the technology has not yet achieved the aspiration.
With all the “AI” hype going on, and the rush to use (often immature) AI tools, my fear is the job of Product Management will skew and morph into what the tools can do vs. what Product Managers should actually be doing.
While many others have spoken and written about the impact of “AI” on Product Management, probably the most visible example was Claire Vo’s talk at Lenny’s Summit in 2024, entitled “Product Management is Dead”. Spoiler: It’s not.
You can also read the blog version of her talk here.
I encourage you to check out Claire’s talk if you haven’t already done so. It’s relatively short and she lays out her argument well, but quite frankly, I don’t agree with her. I don’t want to deconstruct her argument (too much), but I will share the following slides to help you understand her thesis and to dig into why I have concerns about the current thinking of AI and it’s impact on Product Management.
First of all, I think we need to get clarity on what the “talent stack” is, and what parts of it will “collapse”, if that is even the correct word. This is important because ANY hypothesis, and Claire’s view is a hypothesis, needs to be specific on such a key point. What is the entirety of the “talent stack” that is required for Product Management? This is a very important question to answer as it lies at the heart of a lot of discourse about Product Management. Claire doesn’t define it in her talk.
Here we have the Product Triad — a good practice of having Product, Design and Engineering work together on various activities to create common context and stay aligned. But this triad is NOT the entirety of the work Product Management does, and this is where a major problem lies in much of our discourse on Product Management.
Claire sees a merging of these roles because the initial AI tools that are coming out are providing capabilities across these functions, though I would say mostly across Design and Engineering.
I’m not sure what a “Generalist Specialist” is, but yes, there are *some* people who *could* do all of these things — (basic) Engineering, Design, Product — but is it a good idea? And what is the work that is part of that “Product” circle, because, it’s never actually identified.
Then Claire talks about the “AI-powered triple threat” as illustrated above. Quite frankly, this is pure speculation and it’s unclear what this “triple threat” would actually do? Are three jobs going to be done superficially by one person?
There’s so much assumption and ambiguity built into this diagram, both on the AI side of things. (i.e. that “AI” will enable people to do all 3 of these jobs with proficiency.) and on the responsibility side (what are all these roles actually responsible for?)
Finally, Claire brings up the role that Product Leaders will have to play, managing people who do all of these things, apprently including Marketing, Sales etc. This is NOT going to happen. Let me just say that bluntly. I have no idea how this would happen or who would want this.
What problem are we solving?
When I look at Claire’s presentation, this is the question that comes to mind.
What problem are we solving here?
To me, this “AI triple threat” is not solving any problems. It is a solution looking for a problem. And in fact, it is reflective of a big problem in the Product community as a whole. We get enamoured with technology and focus on what the technology can do, vs. what the market needs are, and what important or valuable problems need to be solved.
So, instead of doing what is all too frequently done with products and technology — putting the technology before the problem — let’s start with the problem(s), and then identify how technology, in this case “AI”, and the shifts that could happen, can help; you know, like good Product Managers should be doing. 😃
What are the problems we should be solving?
We know technology Product Management has problems. If that weren’t the case, then there wouldn’t be so much burnout and dissatisfaction amongst our ranks.
I wrote about some of the problems I see in this article below, but I will list out many of the common issues that could and should be addressed to improve our profession overall.
The 5 dysfunctions I listed are:
- Poor Job Definitions
- Under-skilled Product Managers
- Poor Processes
- Unclear Objectives
- Weak Product Leadership
There are certainly more, but those were the ones I focused on because I see them most often.
Thinking about systems of work, it’s hard not to think about Deming. I’m reminded of one of his most famous quotes:
A bad system will defeat a good person everytime.
And, quite frankly, there are a lot of bad product management systems in place in companies defeating a lot of good people. You might be working in one now. Those systems were not designed with intention, but evolved over time, probably reactionary in approach, and likely by several different leaders, some of whom weren’t product people at all.
Now imagine if the Sales processes, Marketing systems and Engineering orgs were “designed” the same way. How successful would those organizations be?
The problems in Product — reflected by burnout and dissatisfaciton — don’t just impact Product Managers. Because product management itself is cross-functional, the product processes touch many other roles. Here’s some data from May 2025 in Lenny’s newsletter, showing burnout rates across different product related roles.
I don’t see this as a coincidence. Design, Research, Growth, Marketing and Product are all tightly bound together and their level of burnout is IMHO basically the same. i.e. about 1/4 of respondents said they were very or completely burned out.
So how bad are the systems? In no particular order, some of the problems I see or hear about in Product Management that should be addressed are:
- Lack of clear (or consistent) definition of the role and responsibility of Product Management
- Lack of clear and measurable objectives
- Poor or ill-defined product processes (or lack of trust in those processes)
- Weak connection between business goals and product plans
- Overfocus on delivery of features
- Lack of focus on strategy and differentiation
- Poor prioritization discipline
- Insufficient user/market research and insight (i.e. weak foundation for decisions)
- Lack of necessary skills/abilities amongst Product Managers
- Weak product leadership
Have I missed anything major?
Does “AI” help with any of this?
Looking at the list above, “AI” won’t automagically fix any of those problems. It’s not like gen-AI can address role definitions or bad objectives or lack of strategy etc.
BUT, and there is a big BUT 😄, the focus on what “AI” COULD do, is opening the Overton Window to making changes in Product Management.
There is an opportunity to FIX many of the problems we’re living with in Product Management, and THEN leverage “AI” where appropriate to enhance how we do our work.
i.e. let’s not simply glom AI on top of the current systems, or let the tech bros drive the changes.
The question is whether we want to be passive and let others (e.g software vendors focussing on low hanging fruit) decide how we will change, or do we want to take the lead, and drive those changes ourselves, to benefit ourselves and the companies we work in?
The negative impact of Agile on Product Management
We’ve already seen what happens when we let others drive changes in Product Management. Yes, I’m talking about Agile. Fifteen to twenty years ago, as Agile was growing in popularity, we started to hear Product Management influencers talking about “Agile Product Management” and things like “Agile Roadmaps”. There were even job listings for “Agile Product Managers”. I remember shaking my head many times as I saw those posts. All of that did great harm to Product Management. I wrote the following— originally in 2008 — on the topic of Agile and Product Management.
Unfortunately, the Agile train, in all it’s flavours, took over the product development world and delivery focused methodologies like Scrum turned Product Managers into Delivery Managers — over focused on feature delivery with little agency and focus on strategy and connections to business.
Bad mantras like “Always be Shipping”, “Move Fast and Break Things”, and ideas like “You won’t learn anything until you ship something” dominated discourse. Developers love these, but for successful products, they were really bad advice.
I’m happy that people are realizing that hiring Product Managers but operating them as Delivery Managers is a bad idea. But now with “AI”, we’re facing another hurdle: the pressure to use and embed “AI” throughout the work we do.
But adding “AI” into broken systems is NOT going to make things better. If we were building a software product, we wouldn’t just add some “AI” onto the existing software. We’d look at the system holistically and understand the benefits that “AI” could provide — the workflows, use cases etc. — and then we’d probably look at the architecture of the software, decide what changes were needed because old assumptions needed to reworked. We need to do the same with how we work in Product Management.
Product Leaders Must Step Up
The change in Product Management MUST be driven by Product Leaders. One of my favourite quotes in this context is from David Marquet.
“As individuals, we should embrace our responsibility for being the best we can be within the design of the organization.
But as leaders, our responsibility is to design the organization so that the individuals can be the best versions of themselves.” — David Marquet
Individual Product Managers are trying to be their best selves, but in fundamentally broken systems. That’s the root cause of the frustration and burnout that we see.
How to (re)DESIGN your org
But it’s up to the leaders to (re)DESIGN the organizations so that the individuals can thrive and excel. I’ve emphasized the word DESIGN because design is intentional, thoughtful and purposeful. And we need that more than ever in how we design Product organizations, the goals they have, how they are organized, the skills people have, and the ways the work is done to achieve those goals.
Without going into it too deeply here, I recommend the following steps. This is a simplified version of the work I do with organizations who are looking to change and improve their outcomes.
- Conduct an assessment of the organization
- Share the findings with the leadership team
- Agree on what the desired end-state should be
- Identify any skills gaps in the team
- Identify IF “AI” can assist or support any of the changes
- Make the changes over time
1. Conduct an assessment of the organization
Speak to individuals to understand what is working well, what isn’t working well, where the biggest pain points and disconnects are. I look at roles, actions, skills, gaps, goals/outcomes etc. and get feedback from Product, but also Sales, Marketing, Operations, Engineering, Design, Support etc.
This takes time, but it’s critical to understand the actual situation. And it’s not always what people think it is. The cross-functional analysis reveals a lot of issues under the surface that few, if any, truly appreciate.
2. Share the findings with the leadership team
From the assessment, share the findings. It will always be a surprise to some folks, as everyone has their own personal slice of reality in the company, as well as their own biases and their own fiefdoms they want to protect. The findings include both the good and the bad.
3. Agree on what the desired end-state should be
This is critical. It’s very easy to just jump into “fixing problems”, but it’s more important to first define the success criteria. i.e. before spending time/money and making changes that impact various teams, get agreement on what outcomes people can expect after the changes are complete.
This will not only help with alignment on the next steps, but you can always reference those goals or that ideal end-state when making decisions or explaining why certain changes need to be made. This also keeps you focused, because it’s easy to get side-tracked during the change process.
4. Identify any skills gaps in the team
People talk a LOT about achieving outcomes or lasting impact. And of course the mantra — Outcomes over outputs — gets repeated far too often without really understanding what it means.
In reality you need outputs (of some kind) in order to achieve outcomes. And in order to get outputs you need actions. And to perform those actions you need skills and abilities.
Understanding the skills gap in your teams and deciding how to address them is important up front. For example, if your team is not great at qualitative research (i.e. defining and conducting customer interviews, analyzing them, extracting insights, leveraging that to drive new products etc.) then getting them “AI” tools to help them with research will be of little value.
5. Identify IF “AI” can assist or support any of the changes
This is the time to start thinking about where “AI” can help in the work that needs to be done. For example, can “AI” help with strategy? If so, how? What tools are there, or information is needed? What kind of process would be required etc?
To be frank, based on my own understanding of AI tools in the market — and I’m not an expert of what is available — what I’ve seen is that some of the more mundane tasks — writing requirements, summarizing or cursory analysis of discovery/research etc. — are common use cases and supported by existing tools. And certainly IF (and it’s a big IF) you want your Product Managers to be creating prototypes, landing pages or mockups of software, that is doable with a variety of tools.
But I ask, and you should ask as well, whether that is the MOST productive use of their time? Are there other people like Designers or Developers who PMs could work with to create mockups, prototypes etc. Do you want your PMs working on mockups/prototypes, or on market/customer research, business alignment, pricing, packaging another tasks distinct from actually building product?
Decide where the Product Managers should focus, and give them the skills and knowledge on how to do those tasks? Design the processes that can effectively and efficiently get the outputs you want. Work with your teams to understand and internalize that knowledge and those processes.
6. Make the changes over time
Organizational change takes time and needs to be measured by relevant proxies. This is the most difficult part. These changes need to happen while concurrently working on what needs to be done. i.e. you’re fixing the ship while the ship is in motion.
Keep focused on the goals and success criteria you set (Step 3 above), but keep assessing where changes will have impact. If you think of your org as a product, then apply the same product principles you use every day to this task as well. e.g.
- Measure what is important.
- Make evidence based decisions.
- Manage stakeholders accordingly.
- Create feedback loops to monitor progress.
- Focus on the long term while managing the near term.
“AI” is a technology. It is not some kind of magical elixer. It won’t solve any problems we have on it’s own. And if we let others decide what benefits it will provide, then we are handing our agency over to others.
Product Management is too important a role to leave to others to “fix”. We can provide huge benefits to our teams and our companies by taking charge and designing great organizations that can deliver the results we need and leverage technology (AI or other) to our benefit.
Are you up for the task?
A Little Feedback Please
If you’ve read this far, thank you. I’d like some feedback on the article to make it better. It’s just a few questions and should take just 1 minute, but will really be valuable to me. Thanks in advance.
===> Click Here <===
