AI may free us from the ‘do it all’ myth — by making it real.
My CTO once looked at me and said, “I’ve never seen a great product manager in action.”
Set aside how that made me feel like my work was unappreciated. Like I was a fraud.
Like all of us, he had been taught that product managers should be able to create a strong vision for a product, work with the team to come up with solutions, make sure the solution got built exactly right, measure it, report back to the rest of the company, and do it for the next item on the list without dropping a beat.
One of the things I took from that conversation: at no point in time did he realize that his expectation of great product management was an impossibility.
There’s no way to be amazing at all of those things, all of the time. The “do it all” product manager is a myth.
Teams that have realized “do it all” is a myth have frequently brought in product ops. With the right support structures in place, great PMs can get closer to living the legend.
Yet even with the best product ops team providing that support, it’s still been a very hard bar to meet.
With the emergence of AI, the conversation has changed yet again.
We now hear the hot takes about the end of product management; too many of us have been part of massive product management layoffs.
Instead, I have a different take: because of AI, we’re going to get several steps closer to the “do it all” PM.
The best product managers have learned what not to do
Since the “do it all” PM was always a myth, the best PMs figured out how to work around it. They selectively chose which areas to excel at, and let others drop.
In some orgs, that would look like a very execution-heavy role, where product was making sure that something – anything – got shipped.
In other companies, the product manager doubled down on strategy and stakeholder management, trusting teams to get something out the door.
Visually, Petra Wille’s PMwheel framework has always served as a good illustration of this. Every PM will be further out on some of the fronts and closer to the center on others.
We can use this same framework to analyze where AI might help create more effective PMs.
There have been two constraints on the “do it all” myth:
- The skill gap: It’s really hard to develop expertise in so many different areas across design, business, and technology.
- Hours in the day: Even for the product managers who are knowledgable about most or all of these areas, there just aren’t enough hours in a day to execute well across all of them.
I’m not sure how much AI can really close the skills gap. To actually understand and be great at product strategy, to truly empathize with users, to translate new product ideas into finance – these will still be the skills we spend careers developing.
But create more hours in the day? AI is fast. Tasks that used to take us hours or days can be done in seconds or minutes. Research, analysis, and writing have all been put on hyperspeed.
I predict that AI’s impact on product management will be that we finally meet the expectations put upon us many years ago. We have a fighting chance to be the “do it all” PM, because we have the tools to speed ourselves up. By walking through each point of the PMwheel with an AI-specific lens, we’ll be able to quantify how much AI has potential to help us become SuperPMs.
How AI enables the Super-PM
There will be some elements of PM work that will be better enabled with AI, some that will get harder, and some that will be reasonably unaffected.
On top of that, product ops teams have a new challenge on their plates: helping PMs figure out how to actually use AI effectively. Not just throwing tools at people, but thoughtfully considering where AI makes sense and where it might do more harm than good.
Let’s break down each piece of the PMwheel and look at where AI could supercharge PM work – and where product ops needs to step in to make sure we’re using it right:
| Dimension of Product | Will AI Speed Product Management Up? | Product Operations’ Role |
| Understand the problem | ✅ Yes – AI accelerates synthesis of interviews and analytics, reducing time to insights | Make research data AI-readable; vet and implement AI tools that aid but don’t replace human insight |
| Find a solution | ✅ Yes – Vibe coding, prototyping, and ideation are faster with AI | Curate design systems/templates for AI tools; reduce friction for PMs to prototype quickly |
| Do some planning | ⚠️ Somewhat – AI can help think through options but doesn’t replace strategic thinking | Traditional ops work still most relevant; AI’s role here is limited and better addressed elsewhere |
| Get it done | ✅/⚠️ Yes, but… – Faster eng velocity increases delivery, but burdens PMs; AI helps with prototyping and coordination tools | Unclear value right now; maybe tooling support, but likely not a primary investment area |
| Listen and learn | ✅ Yes – AI can synthesize data into dashboards, making it faster to understand how a product is doing after launch | Build systems that track performance of shipped work and surface learnings quickly |
| Team | ❌ No – Relationship-building and influence are human activities | None; this is human-to-human work, outside AI’s domain |
| Grow | ❌/⚠️ No, or slower – Learning AI is a new demand; AI doesn’t reduce learning effort for PM skill development | Support self-guided, AI-powered learning paths; enable structured growth journeys but expect time investment |
| Agile | ⚠️ Indirectly – Faster prototyping supports agility; AI enables speed but not mindset | Build systems that allow fast releases and experiments; help teams leverage AI within iterative workflows |
Understand the problem
Is the product manager aware of the underlying user problems of the product they are working on? Does he understand the motives, issues, beliefs of these people? And has he thought about what the needs of the company/organization are when it comes to creating this product?
AI has already started speeding up data synthesis and interview analysis. But, from product managers I’ve coached, the biggest blocker to understanding the problem is time to talk to the customers.
In terms of how product ops should be approaching this area, there are a few ways to be high-impact. First is that all the research data – interviews, insight reports, and analytics – need to be easily accessible by AI. Ops should be thinking carefully about how to get that data to be available for AI to read through.
Second is diving into AI-powered tools to assist with this. This is one area where I’ve seen some really great products on the market already. They don’t replace the insights we build as product managers, but are great at helping better validate the intuition we’re building as we explore the problem space.
Find a solution
Did she find some good problems to solve? Great! Can she come up with some possible solutions and experiments for testing them?
If there’s one thing that AI does great, it’s creating fast prototypes to help come up with a good idea around a solution. Vibe coding should allow us to try out more ideas, get them in front of users faster, and figure out what direction to go.
This is happening already for many product managers. Where product ops can help enable this more is around getting design systems, templates, and other resources into the coding tools so product managers don’t have to set up their vibe coding environments in order to get goign.
Do some planning
No matter if you are a fan of good old roadmaps, or you know the latest Agile planning tricks, a PM must have a plan and a story to explain what’s next.
AI will be less helpful here because planning really means creating a cohesive strategy. AI can be a thought partner and speed things up a little, but won’t be able to turn hours into minutes or seconds.
Meanwhile, there’s lots that product ops can do to enable better planning at the PM level, but very little of it is going to be AI-driven. As a result, I’ll save that for a different article.
Get it done
Every PM needs to know how to work with her product development team to get the product out to the customer.
AI’s impact in engineering has been transformative, with trickle-down effects into product management. As engineers code faster, that means more things getting shipped, more product going out the door, and more getting done.
At the same time, a faster engineering velocity means more for a product manager to keep up with. We’re seeing a lot of companies trying to get AI ticket writing or project management going and claiming that that is going to be replacing product management. As well, the vibe-coded, stronger prototypes are going to make a big difference in this environment because it’s going to create stronger levels of communication between product manager and engineering team.
Yet as we all know, more outputs does not necessarily mean more outcomes. If we reduce the number of product managers while increasing engineering velocity, it’s likely that we’re going to end up with products that have a lot of features and not a lot of value.
Product operations potential in this space is less clear. There might be something to be done with tooling, but in general, this is not an area where I would see product ops wanting to invest more of their energy at the moment.
Listen and learn
Once you have released something new, you will want to observe if and how people are using it and iterate on the learnings to improve the current status.
On top of that, if you’re releasing a lot more, that puts a much bigger burden on the product management team to be monitoring release features for performance.
Eventually, AI will be monitoring all product releases based on the desired outcomes. Not just creating dashboards, but rather creating comprehensive reports that tie together customer support, sales data, usage analytics, product analytics, and usage information into a cohesive point of view on how any product that’s in the wild is performing.
Product ops has tremendous leverage here to speed up the work of a product team. Following up on released items is one of the most neglected parts of product management, largely because organizations have a habit of ship it and forget it. But if AI makes it so we can’t forget it after we ship it, perhaps we’ll see more going back to our features to iterate (see also: agile).
I look forward to the ways in which product ops dramatically accelerates the listen and learn cycle in product development.
Team
How good is the PM when it comes to teamwork? What does he know about lateral leadership and motivation of teams?
Team-building by definition is about human relationships, and so has low potential to be changed by AI.
This is the kind of work that we have to do as people: creating relationships, getting to know others, and building trust over time.
Grow
Is the PM investing some time in their personal growth as a product person?
Grow is an interesting one because I think that there are two angles to this:
First, actually learning how to use all of these AI tools. That’s not going to be something where people are moving faster because it is something where it actually is going to be taking up more time – we are learning a new skill.
Second, just because we’re learning about AI doesn’t mean we can stop learning about all of the other skills that are required for product management.
From a product ops perspective, I think that there are going to be ways that we can enable our team members to grow more easily and perhaps use some of the newer tools coming out that have customized learning paths. But ultimately, this is an area that will take up a lot of time post-AI and probably will continue to do so because a computer cannot, at the end of the day, help us learn or develop that much faster.
Agile
Is she just living in the Agile world or do they fully understand agile values, principles, and ways of working?
This skill is about true agile– building quickly and releasing iteratively.
The relationship between AI and agile is complex. While AI dramatically increases our shipping speed, which could enable faster iteration, it also makes it temptingly easy to overbuild features. Product teams will need to actively resist the urge to pack more functionality into each release and instead maintain their commitment to small, iterative launches.
Product ops role in agile becomes more interesting because I believe that there might be more cases where product ops teams are building internal tooling to enable these fast iteration, release, and feedback cycles.
Product ops won’t be replaced by AI
Imagine a world where product managers are able to live up to the expectations put upon them.
It will not mean the death of product management or product ops. Neither will it lead to a massive increase in the roles.
Product Ops will be a very effective way to super PMs to exist. Highly AI-leveraged employees do not magically appear; they need infrastructure to exist.
And, just like product management, product operations is going to be expected to be able to do more and to do it faster. So we’ll also need to find ways to leverage AI and individual workflows in order to be more effective at delivering team-level value.
If you’re trying to navigate the world of AI-enabled product operations, I am looking for some design partners on an AI enablement sprint product that I am building. Please reach out and we can see if there’s a good project where we can work together.
