The culture of experimentation where data trumps intuition and a system that makes running experiments easy allow quick and effective innovation.1
Speed of iteration beats quality of iteration. (Boyd’s law of iteration)1
- [+] Faster you can get something in front of users, faster you’ll verify you are building what users want and adjust the course.
- [+] To achieve faster iterations you must strip all unnecessary requirements from scope and you’ll really be developing a MVP with less over-engineering.
- [-] MVPs can be used as cover for rushing to a less thoughtful solution without considering the context or the long-term implications.2
- [-] If the quality is bellow a certain threshold, users won’t use the product and you’ll get no useful feedback except “The app is broken!!”.
When you’re working on end-user software, you should design the UI first.1
- [>] Change in requirements will often happen in spite the final version of UI was agreed on.
- [+] A lot of edge cases can be discovered earlier in development cycle (when no code that would need refactoring has been written).
- [+] Stuff that’s not possible to easily implement in UI can be discovered and possibly excluded from the project if not critical.
- [+] Much easier to create a “contract” with product managers of what will be implemented and what won’t - less “feature creep”.
- [+] When coding you know “exactly” what kind of UI needs to be created and what should the API look like to support such UI.
- [+] Less blockers because of trying to coordinate a UI change with product managers because an edge-case was discovered while coding.
- [-/+] Developer takes on more responsibility if the end product is bad.
Stop thinking like “Homo Logicus” and start thinking like “Homo Sapiens” when developing end-user software. Developers sometimes can’t comprehend that the average user doesn't even know what ALT+TAB does.1
- [>] Expose developers to users using the product to make this even more apparent.
The huge body of intermediate users is so dominant that you can and should ignore both beginner and expert users.1
- [+] Development takes less time.
- [+] Product’s feature set is focused on doing what really matters.
- [-] If new users aren’t gracefully introduced to all the intermediate features, they may find the product too complicated to use and leave.
- [?] Expert users may bring a huge pile of money which you’ll pass on while ignoring them.
It’s easy to get lost in the details when making decisions about complex systems - it’s comforting to focus on getting the details right while ignoring the larger implications.2
Observe how users actually behave versus the way they tell you they behave/will behave.1
- [>] What users say they will do, and what they actually do, are often two very different things.1
- [>] "Feature fatigue" -> users want as much features in a product as possible until they get to use the product and become overwhelmed by all of these features.1
- [-] Marketing would want to have more features because of “feature matrices” since they are important when trying to look better then competitors.
People are the source of, and the solution to, all the problems you’ll run into when building social software.1
- [>] Create a way to collect feedback from your users and carefully evaluate it.