You Should Probably Stop Running Product

Founders know they have to give things up as the company grows. Sales gets a leader. Engineering gets a leader. Recruiting becomes a function. Operations gets built out.

Product is different. Founders hold onto it longer than almost anything else because they are still good at it. That is what makes the problem hard to see.

At a company doing roughly $1M in revenue, the CEO still works directly with engineering. Product ships on spec and on time. By the standards most founders use, nothing is broken. But the company cannot describe its ICP. Not "hasn't refined it." Cannot describe it.

Growth has come through founder-led sales, which is what hides the problem. The founder is the customer research function — knowing why customers buy, which objections matter, which use cases are real and which are edge cases. Most of that knowledge lives in the founder's head. That works while the founder is doing the selling. It stops working when the company needs a team to sell. A sales team needs an ICP that exists outside the founder's intuition, and product needs the same thing: a clear understanding of which customer it is being built for, and which of that customer's problems matter most. Shipping on time is not evidence that you're building the right thing.

At a company doing roughly $10M, there is a capable product team, but the CEO still sets the product vision and drives the roadmap. The CEO is very good at it. Again, that's the problem.

The question isn't whether the CEO can keep doing product well. It's whether product is the highest-value use of the one person who can do the CEO job. I think of that job as four things: set the vision, hire the team, raise the capital, realize the value. Almost everything else can eventually be done by someone else.

So the comparison that matters isn't whether the CEO is better at product than a product leader would be. It's the value of another hour spent on product against another hour spent on something only the CEO can do. Nobody walks into the CEO's office and says the financing started three months late because he spent the quarter debating the roadmap. The connection is indirect, which is why it gets missed. The product work stays excellent. The CEO work is what gets crowded out.

At a company doing roughly $100M, the CEO is no longer running product — and neither is anyone else. There is no formal product leader. Product exists across technology, marketing, sales and the leadership team.

That works for a while. Technology knows what can be built. Marketing knows how the company wants to position it. Sales knows what customers are asking for. Leadership knows the business priorities. When the product is relatively simple, those perspectives can be reconciled in conversation.

It gets much harder when what engineering ships is not a feature but a platform. A feature can often be adopted within an existing workflow. A platform changes the workflow itself. Roles change. Teams have to understand what the platform enables, when to use it, how to sell it, how to implement it, and how their own jobs change around it.

Engineering can build and release the platform. It cannot own that organizational translation. Marketing can help communicate it. Enablement can train against it. Operations can redesign workflows. Sales can work it into the commercial motion. But someone has to understand the product deeply enough to lead the translation from technical capability to customer value, and then into how the organization actually works. That is product.

The absence of that ownership often shows up first as argument. Engineering and marketing fight over release timing. Sales and product fight over roadmap priorities. Someone announces a capability before it's stable, or refuses to announce it long after it's usable.

The instinct is to call this a people problem. Often it isn't. Each function is doing exactly what it should against its own objective. Engineering is optimizing for technical integrity. Marketing is optimizing for the commitment it made to the market. Sales is optimizing for the deal in front of it. None of them is necessarily wrong. Nobody owns the tradeoff, so the same tradeoff gets litigated again and again in different meetings.

Recurring arguments among competent people are often evidence of missing ownership, not broken relationships.

That is what founders misunderstand about product. They define the function by its activities: managing developers, maintaining a roadmap, writing requirements. The role is to sit between what the customer needs, what the business needs and what the technology can support — and to own the tradeoffs among them.

The shape of that role changes with the company. Early on, the founder may perform it legitimately. Later, a dedicated product manager may be enough. Eventually, it requires a product leader with enough authority to decide across functions.

The transition happens later than it should, and not because founders are bad at delegating. They give up the things they are bad at first. Product is often one of the things founders are unusually good at. Competence at the second job disguises the cost of continuing to do it.

That is a specific case of a broader pattern. As companies grow, roles get smaller. Early on, one person does five jobs because that is the cheapest way to coordinate five jobs. As complexity increases, keeping all five inside the same person's head stops being efficient. Ownership has to get clearer. Functions specialize. Jobs narrow.

The CEO's job narrows most of all.

The question isn't whether you're still good at product. It's which CEO job you'll be doing badly twelve months from now because you're still doing this one.