“You know the real difference between me and a believer? It strikes me that religion, in its essence, seeks to take natural events and ascribe supernatural causes to them. I, however, seek to take supernatural events and find the natural meanings behind them. Perhaps that is the final dividing line between science and religion.”
- Jasnah Kholin, The Way of Kings
I recently struggled to explain to my PM exactly why the underlying approach I’d been seeing feels so wrong. The quote above really struck me as analogous. Here is the example:
The PM would get some flavor, some feel, of what we should do next. Then he sells it hard. When on calls with customers he pulls them by the nose directly to a specific answer he is looking for. Then he has the power to say to anyone in the company, “I talked to the customers, this is what they said”. Technically true, but not truthful. So he goes of selling hard internally and externally about this great new world we’re going to build. None of this, by the way, is talked over with engineers who will be building the product.
I’m pretty used to ignoring this kind of warped marketing/sales behavior, but what got to me is the ending. He booked a big launch demo event. To be clear, we didn’t talk about or scope the work or even really decided on what we wanted to do, and he booked a demo event. This is bonkers.
So how should it work. Returning to the quote above, what the PM here was doing is taking a nothing burger and making it this magical fix, as if just talking about it enough was enough to bring it to reality in the world. What we should do is the opposite. We should first build. Put something into the hands of real customers and see what happens. Something small and not quite finished, but something that will give us real data of what real customers will actually do. Then we should try as hard as we can to not fool ourselves. We should be our own worst critics poking holes in our design in every way we can to make it better based on what real customers are doing. Then we should refine or throw that feature/product away based on the data, not holding onto to any one idea or set of ideas too strongly, and working hard to not invest too much time and effort into something that is likely to fail. An it is likely to fail, at least for the definition of failure we should adopt.
Another analogy came to mind: would you rather have a super fancy sports car with a crap engine in it, or a normal looking sedan with a race car engine? Product organizations who take the sell first approach are building the super flashy car with nothing inside. This may work to sell a few cars, and I’d even say this is fine as long as you’re truthful about what the car really is. But don’t promise it will win races. If you’re looking to win the race and really build the best product, always build first and sell once you know you have something.