When I designed laser chips for optical networks, we always had a general specification to work from. Being young, I imagined that these product sheets were dreamed up my geniuses with crystal balls. Turns out they were written by product managers.
How did the product manage come up with the specs? They talked to our customers. They asked about their problems. They came back to the mother ship and asked us, the designers, what could be done. They then went back to the customer with some trial specs. The customer would take that backs to their designers for feedback. This would continue until there existed a product sheet that we thought we could build and the customer thought would be useful.
Now this wasn’t work for hire. This wasn’t a contract to produce ten thousand parts. This was simply an understanding that “if we could produce this produce it might be useful enough that our customer would buy it.” A minimum viable product sheet.
Starting my new company, we didn’t start with a fully built product. Or even a web site. We started with a PDF listing bullet points we thought might be useful to our potential customers. We got them on the phone, we asked them to review the bullet points. We iterated until we had a list that people were saying “yeah, that would be useful.” Then we started building.
I’m not going to say that this is revolutionary or that it pertains to startups or software. After all, it is exactly when we were doing in a 100 year old manufacturing company. I think that in software in general, and web in particular, we skip the minimum viable product sheet and go straight to the minimum viable product. Perhaps if we slowed down, started with the product sheet first, our products might just be that much more awesome!