I recently run a poll amongst managers to check on what the priorities are for them.
Here are the results and my comments:
New features (9 votes) vs better existing features (6 votes)
As you can see they are very close.
I read this as more or less “let’s not add new features fast just to have them, but let’s make the system stronger every day and also add new features”.
This may seem obvious but it’s not.
It goes down very well with the decision to develop incrementally, that is: built a new feature very basically and then iterate (improve, get feedback, improve, get feedback …) quickly.
In other words, each time you get a new feature it will be a half baked product good enough to let you do very basic things.
But since basic things are 90% of real world usage it makes sense.
Often a new feature will generate a lot of work for you, so you tell us “come on, make this automatic” and if your request is useful also for most users, we’ll do it.
The best thing of this approach is that we can avoid the sexy but unuseful functionalities, the ideas you get when you develop lost in your nerdy world with absolutely no connection to reality.
Those normally come out like “let’s do a calendar which auto generates rss and sends an email to every Owner except those who have a conversion rate lower than 15% and red hair, and let them order pizza online without even needing to know it because our algorithm decided they are hungry“. Continue reading Priorities