Skip to content

Will We Finally See the End of the Product vs. Project Debate with AI?

Stream Spy vs. Spy (prod. Windxws) by Mr. Murk | Listen online for free on  SoundCloud

Since “the beginning of time”, there has been a discussion in the software world. The debate between product and project management has been a staple challenge in the software development world. We have entered a phase of radical technology innovation, driven by the rapid advancement of artificial intelligence, we may finally see a shift in this dynamic. I believe AI has the potential to refine these roles and redefine how we approach software development.

The Heart of the Debate

Before we get into it, let me be clear.  As a product leader, I have nothing against project management.  Good project management is an essential need in a huge swath of companies in and out of the technology realm. Also, let the record show my wife is a PMP certified project manager, so yeah… nothing against project management. 

At its core, the product vs. project debate revolves around the fundamental responsibilities of each role. Marty Kagan perfectly describes product management as the role responsible for delivering Value and Viability. He describes Value as the external measure, and Viability as the internal measure.  I wholeheartedly agree with this perspective. Product management is about ensuring we're building the right thing for the customer—a solution that delivers value and can be successfully delivered.

In contrast, project management focuses on the efficient and predictable execution of a project from start to finish. It's about managing timelines, resources, and deliverables to ensure the project is completed on time and within budget.

The Challenge of Empowerment

The challenge arises when organizations struggle to truly empower product managers to fulfill their role of delivering value and viability. Too often, product management is pushed into project management, forced to prioritize roadmaps and fit product development into a measurable and predictable box.

As this happens, the product development process starts to resemble a project. Features are mapped out, timelines are set, and the entire organization aligns behind a neatly packaged roadmap. The problem with this approach is that unless you're building something already built before, that feature map is either pure fiction or a path to mediocrity at best.

The Uncertainty of Innovation

The whole point of being an innovative startup is to build something new, solve problems that haven't been solved before, or at least solve them in novel ways. When product management is forced into a project management mold, it stifles innovation and leads to a feature factory mentality. Product managers start looking at the path of uncertainty ahead and mapping out the things that seem like the surest bet from where they sit right now.  That often means something that is a variation of what already exists in the world.  The more certainty of delivery timing is demanded by the leadership team, the less likely the product org will deliver true innovation.  

In the short term, the organization may be happy with the predictability and alignment, but in the long run, disappointment sets in as the product fails to deliver true value and differentiation.

The Creation Changes the Creator

Enter artificial intelligence. The introduction of AI into the software development process will change everything. We're already seeing this with tools like GitHub Copilot, which are writing more and more code. I believe we are just scratching the surface on this.  In time, I believe most actual programming will shift to AI.

This doesn't mean developers are out of jobs. Rather, it means they can be true engineers, adding value where they're needed most. Engineers don’t just build things.  They solve problems. They develop the best possible solution to a given problem that meets the constraints provided to them.  In fact, outside the software world, this is ALL an engineer does.  Civil Engineers don’t build the actual bridge.  Electrical Engineers don’t build the production circuit board. In the future, I’m willing to bet that software engineers won’t build the actual software.   

Along with this, product management will need to shift in response. 

Sidenote: Interestingly enough, developers tend to get marginalized into programming roles in the same organizations where PMs get marginalized into project roles.  It goes hand in hand. Alas, that's a whole post in and of itself

The Evolving Role of Product Management

As AI reshapes how software is built and shifts the focus from task-oriented UIs to goal-oriented solutions, product management will need to evolve. 

What does that look like? A few potentials 

  • PMs won’t be writing specs … the AI can write these if pointed in the right direction
  • PM’s won’t make estimates … AI will be vastly better and more objective at this 
  • PM’s won’t need to map out a roadmap … AI will be better positioned to do this with a data driven real time update approach
  • PMs won’t need to do QA or Bug triage … AI will be inevitably better at this 

If we remove these busy work and project tasks, what’s left? How does a PM add value? Maybe we don’t need PMs or product at all? 

In truth, the entire process of product development will likely increase in velocity with machines writing code guided by engineers. Testing happening by machines.  Process plans mapped out by the AI. We will likely have substantially more capacity to build more at a pace we’ve never seen before. Who will decide what problems to point this new high powered production machine towards?  We’re likely going to need true Product Leadership more than ever.

The nuances of getting things built on time will likely fade, and the real value will lie in choosing the right problems to solve. The creativity and strategic thought that goes into thinking about the problems at hand and identifying which of those if solved would deliver the greatest value to the customer.  Put simply, we will still need PMs to be the steward of the customer and how the business delivers on the promise most effectively. 

Isn't that what we always wanted product managers to do? Yes, indeed.

With AI handling the execution and delivery aspects, product managers will be free to focus on the core questions of Value and Viability. They can dedicate their time and energy to understanding customer needs, identifying opportunities for innovation, and ensuring the product delivers real value.

The Pains of Experience

In my own career, I've faced the challenges of the product vs. project debate head-on. I've had successes in overcoming it and failures in escaping the project trap. Much of what made us successful at HubSpot in the early days was the radical reshaping of our product development organization. Prior to that, reshaping the product process at HubSpot was predominantly sales driven.  What to build was decided for the most part by the executive based on what would drive revenue and be delivered in a predefined time. As a result, the product was a hodge-podge of features that was nothing like the cohesive all in one product you see today.  HubSpot was of course a sales and marketing machine in those days.  It made sense to try and build a product org to match.  It wasn’t until we tore that process down and rebuilt one that completely decoupled the product development org from the go to market motions that we were able to build a product that truly delivered on the promise.  

We empowered engineers to be engineers and own the product they built. We pushed product managers to relinquish the reigns of timing, delivery and project management, and focus on talking to customers and becoming experts in value and viability. The transition was tough for the entire org, but I doubt anyone would look back and do it differently now.  

In another company (who shall remain nameless), I attempted to recreate the magic of HubSpot.  I pushed hard to eliminate roadmaps and commitments other than a commitment to deliver value.  For a while, it worked.  However, as the team grew, and the demands of the board of directors along with it, the system started to shake. Eventually, we started to relent and provide more time-based commitments.  A map of the future that represented our view of the world from where we sat.  It felt clear at the time.  It didn’t feel like a compromise. We celebrated that pace of delivery above all else. Unfortunately, the result was a feature bloated product, with 80% of those features completely unused.  Worse than that, the needle was not moving on the business. We didn’t deliver value. We just delivered on time.  We fell into the project trap. The worst part is, the product team took the blame regardless (as they should).  No matter how many green check marks are on that board slide from the prior month, if the numbers don’t move, the product isn’t where it needs to be.

These experiences have shaped my perspective on the potential for AI to potentially lessen this problem. I believe we're on the cusp of a paradigm shift in software development, one that will redefine the roles of product and project management. As AI takes on more execution and delivery responsibilities, product managers will be empowered to focus on what matters—solving the right problems and delivering value to customers.

The PM vs the PM 

The debate between product and project management has been a long-standing challenge in the software development world. With the rise of AI, we may finally see clarity, at least in terms of where the real value comes from the product org.

As AI reshapes the software development process, product management will evolve to focus on the core questions of Value and Viability. Product managers will be free to dedicate their time and energy to understanding customer needs, identifying opportunities for innovation, and ensuring the product delivers real value.

The future of product management lies in embracing this shift and leveraging AI to drive innovation and deliver truly valuable solutions to customers. It's an exciting time to be a product manager, and I believe those who adapt and evolve with the changing landscape will succeed in this next era of software development.

Latest posts