Culture ate strategy. AI is devouring the 2-pizza team. Now what?
AI tools are changing everything about how Products are developed, companies are run, and how we work together . Except for what it’s not changing at all - the underlying philosophy of Product thinking.
I’ve been having a ton of great conversations with people on the front lines over the past few weeks. Things are changing way too quickly for there to be any answers, but we’re starting to ask the right questions.
Let’s get into it.
What does a product team look like these days? How many people belong on that team?
Amazon introduced the idea of a 2-pizza team - a team small enough to be fed by 2 pizzas, but with the capabilities to be self-sufficient in their mission. Over the years, that usually looked something like 8 people - 1 Product, 1 Designer, 1 Dev Lead, and 5 Devs, with some room for variation.
But now we can hook up a design system to an MCP and one person can prompt their way to production-ready code. Anyone with the right mindset - whether they're from a Product, Design Dev, or any other background - can build good features and ship them. And with the right Agents set up, we can even get feedback from synthetic versions of the other functions. Or even synthetic users.
We lose a lot when we operate that way, though. There’s real value in having your ideas tested and refined through collaboration, and in being exposed to other people’s viewpoints. And I’ve been hearing first-hand from people that they notice real differences in the way that people with different backgrounds and mindsets generate prompts - and what that results in.
There’s also an inordinate amount of key-person risk when everything is centralised into one person’s brain, even if they offloaded or backed up a lot of their thought process into a digital brain.
I don’t know how many people should be on a team these days. Probably less than 8. What matters is the overall capabilities of the team, and what they can accomplish together. My definition of an effective Product team has always been a bit different, and it seems even more relevant these days:
A Product Team is a group of people who can:
Ask a question (and knows which question to ask)
Get an answer
Understand that answer
And make a decision and then take the appropriate action(s) based on that understanding
What does the product development process look like?
The underlying philosophy hasn’t changed much. We’ve been talking about Build / Measure / Learn forever, and that’s just as effective as ever, if not more so.
What has changed is the cost of those phases. We’ve internalised that Discovery comes before anything else to an extreme, forgetting that we can discover things THROUGH the build process - and that made sense when build was expensive and time consuming.
But now it’s cheap, and fast. Building a production-ready system that can scale is still hard, even if it can be faster. But prototypes? Exploring the art of what might be possible? Anyone can do that now, anytime.
What’s critical is being able to do that in the service of the most important opportunities for your organisation. To prioritise the effort of your most valuable resource: time and effort.
There’s no single answer that anyone can give right now. (To be fair, there has never been a single answer, no matter what the books say.) It’s too early. It’s terrifying and exciting, and we need to experiment widely in how to apply the underlying philosophy of Product thinking.
Should there be any product managers at all?
Maybe not. For my entire career, Product people have been asking for a seat at the leadership table. It’s the wrong question. It always was.
No one is ever going to give you a seat at the leadership table. The rest of the leadership team - at least in legacy organisations - has always thought that they could do our job. Technology has always been a thing commissioned by other departments. They talk to customers all the time; they can see the problems and they know - in their guts - what needs to be done to fix it. It’s simple - just do this one thing, and things will improve.
Of course it’s not that simple. It never is.
But we have an opportunity to change things now. It doesn’t take weeks or months to respond to business priorities; we can start to respond to things in days now. And that changes the equation.
If that’s the case, it means that we don’t have to be stuck in the Feature Factory delivery trap. After all, who knows better what needs to be done than someone trained in Product thinking?
It doesn’t mean that we have to sit in the Product Manager’s chair, however. Maybe we should consider how we can be at our most effective in an organisation. Maybe Product Management as a distinct practice in a company was always a feature, not a product. Maybe we should be putting ourselves at the heart of business decisions, wherever that sits - in Strategy, General Management, Finance, or Marketing teams - and effecting change from there.
Where does Product Management fit into an organisation?
This is a problem in lots of companies. In start-ups, it’s often a founder’s job until they get to a certain size. In legacy companies, it often sits under IT. And we’re seeing more and more CPTOs to ensure that product strategy and delivery are aligned.
The reason that Product was established as separate from business units is because it’s new(ish) and needed to be protected. But that’s caused at least as many problems as it’s solved.
If we disperse Product thinking across the organisation, we can separate IT into its component parts: the infrastructure that keeps an organisation running - Communications, Finance, HR, Security, etc - and the new product development elements that often need to move more nimbly, and be closer aligned to realising business value at pace. And if we can change the team’s makeup and start working by building to learn and then refining to deliver…. Well, maybe the latter also sits in a functional business unit and is 100% aligned to business goals.
That’s not to say that there doesn’t need to be a coordination layer - a guild for builders, for example. But this model already exists in many businesses, where Finance sets standards centrally, but places people directly into teams to work as part of them on a day-to-day basis.
There’s plenty of other ways to make this work, of course. And this approach might be a disaster - but it’s a good starting point to open up the discussion.
So, how do we work together now?
It depends. Every organisation is different. People, as Janna Bastow says, are hard. Lots of people, dealing with lots of uncertainty, are exponentially harder.
We can’t stand in the way of progress, or perceived progress. There are plenty of terrifying and terrible things that AI tools will accelerate, but there are also lots of ways that they will be helpful. They are, after all, just tools. It’s how we use them that is key.
As product people, our job has always been to help people make better decisions, faster (quoting Monika Turska). That’s more important than ever now. Having better conversations, understanding the impact of things, balancing strategy, tactics, and ethics to ensure that the right things are prioritised - and realised. Monitoring the risk of shortcuts and the sunk-cost fallacy. Scanning the horizon for change and how it might impact our organisations and strategy.
Everyone else has long thought that they could do the job of a Product Manager. Don’t be defensive about that. Invite them in. Creating reports, roadmaps, and user stories was always only one part of the job. Let go of the things that others can and want to do, and concentrate on new processes, ceremonies, and cadences that work to help everyone succeed.
Now what / So what?
These are going to be hard times. We’re already seeing layoffs coming, and more will likely hit in the near future. That’s going to be extremely difficult for many people. It’ll present a lot of opportunities for some to go off and do new things, to start their own companies or work in different ways.
No matter where you are, the key things are to ensure that you and your teams are delivering value at pace - making an IMPACT (as per Matt LeMay), and that you’re leaning into the change, and helping lead our organisations through it. Start by asking yourself three questions:
What do I need to do to prepare myself for the changes that are coming (or already occurring)?
What does my team need to be successful as this impacts the company and the market?
How can I help my organisation to survive and thrive as change continues to accelerate?
I'm Randy Silver. I work with leaders and product teams navigating the organisational side of AI change — the people, the structure, and the decisions that tools alone can't solve. If any of this resonates, I'd like to hear what you're seeing in your organisation. The first conversation is just that — a conversation.