> Everyone needs to care about the Product

This isn't the first time I hear this, but I always have a bit of trouble with this one. It's one thing to take a step back and think about the actual product and how it'll be used, but I think it's presumptuous to think that software engineers know what makes a product good or not. We don't say "Everyone needs to care about software architecture, even Product", so I'm not sure why we think the flip side of that is true.

If you consider product as a proxy for customer, I think it gets a bit easier to understand. Customers don’t care about architecture (unless you have a technical product where they do actually need to know architecture). They don’t care about many of the details. They just want their problem solved.

For software engineers, our goal isn’t to necessarily know what makes good product or not - but we do need to make sure that what we’re building solves an actual customer problem or need.

> We don't say "Everyone needs to care about software architecture, even Product"

We absolutely should say that. I was an engineer for 13 years and have now been in Product for 8 years. I work on a highly technical Product team, and it is absolutely an expectation for myself, my peers, and my reports that we should ensure we fully understand our Product, including its software architecture, and have an opinion about it. Engineering ultimately decides the "How", but they cannot do that effectively if Product cannot articulate an opinion about the architecture guided by an understanding of things like expected scale, potential future integration decisions, and other cross-organizational expectations that may not yet be codified. In general, Product should have an educated opinion on anything that is a one-way door, and so should Engineering. It should not be a unilateral decision, and if either party is unable to form an informed opinion, that's an organizational miss.

At the end of the day, the goal is to make a product that people find useful. How that ends up happening is almost completely irrelevant to the people actually using the product.