What versus How

Over on the Requirements Defined message board, a guy who always seems to ask the right questions about product requirements is Marc Talbot. In a recent discussion, he questioned just how straightforward the traditional distinction between the what (requirements) and the how (design) is:

I'm Sony.

I want to entertain people (what I want to do).

I'm going to sell a new TV (how I'm going to do that).

Does that make everything related to the TV a design decision?

The What and How are very tightly coupled to the particular problem that you are choosing to solve, and making the decision on what problem to solve is the real trick.

This example very neatly illustrates how difficult it can be to distinguish between requirements and design.

Clearly, there is a huge leap between entertaining people and providing them with a television. The goal of entertaining people is so broad that Sony would want to constrain it. And it's a very important point that choosing the problems to solve determines the requirements. So the idea of entertaining people with a TV seems a reasonable constraint that falls within the realm of requirements.

Yet is it really?

"The product shall entertain its users" may be a functional requirement, but what about all of the associated nonfunctional requirements? What about all of the other problems users are trying to solve or avoid? What motivates them to watch TV instead of experiencing other forms of entertainment? Besides usability, availability, and other standard requirements, what about:

• location - users want the entertainment at home (i.e. going to a concert or play won't cut it)
• realism - users want realism in their entertainment (i.e. audio by itself won't cut it)
• variety - users want to benefit from the various media that exist (i.e. being able to do things like attach a DVD player and watch DVDs is important)

Now, formulating these constraints in comprehensive and measurable terms is a challenge. It sure would be nice if we didn't have to bother gaining an in-depth understanding of why people want TVs. But it's just the sort of challenge the most talented and strategic product managers are ready to face.

Once we've fully understood and documented all of these constraints, a TV likely will be the ideal solution. But it might be a very innovative form of TV - so innovative that it would be the first in an entirely new product category. This sort of innovation starts with examining the true underlying requirements rather than assuming an established product category and its me-too feature set.

[This response is cross-posted on the message board.]