On a high level, I find there are two types of applications you find on iPhones (and, I guess, Android devices): discrete and continuous. This breakdown is by usage profile. You use discrete apps infrequently and to accomplish a certain objective.1 You use continuous apps frequently, you may use them to accomplish a certain objective but you also use them passively (maybe even… aimlessly).
App categories that I would throw into the “discrete” bucket include food, transportation, and finance. Such apps include Yelp, Foursquare, Uber, CreditKarma, etc. App categories that I would throw into the “continuous” bucket include music and social media. Such apps include Apple Music, Facebook, Snapchat, Instagram, etc.
There’s definitely some grey area between the two but the purpose of this thought montage is not to provide a clear CB Insights-esque map of the types of apps or write a McKinsey-grade report on the mobile app ecosystem. It’s only to get some thoughts out of my head of the requisities for discrete apps.
Discrete apps, in my humble opinion, do not serve to entertain. They serve to accomplish a goal or objective. And that goal or objective is sought after infrequently. Infrequently could mean less than every other hour. More often, less than daily and many times less than weekly.
These apps need to accomplish 2 things in order to not be guaranteed failures. These are requisites for being not dead, not necessarily prescriptions for how to be alive.2
They must provide a lot of value.
This isn’t a controversial statement. The typical Silicon Valley sound bites say that any startup must provide a 10x improvement over the current status (whether that’s price, quality, quantity, feature set, etc). However, the level of value required for a discrete app to not be dead is even more important to consider.
It is important because of the frequency of use. A discrete app is not often in a user’s daily routine. Such an app is not often in a user’s working memory. It occupies a small space in its users’ heads, if at all. In order to be called upon by whatever trigger - the need to find a place to eat, the need to get to a spot across town - the app must cement its value in relation to a certain action or need.
In order to do so, it must provide a lot of it. Lest it immediately be forgotten and the user reverts to his or her current solution.
For those fond of equations, treat the probabiltiy of success as some function of
marginal improvement per engagement X # of engagements. The fewer the
# of engagements, the greater the
marginal improvement per engagementrequired to yield the same probability of success.
They must do so extraordinarily quickly.3
This is a related point but equally important to consider. Because the app typically serves one purpose, the barrier to achieving the goal must be extraordinarily low and the goal must be achievable within a quick turnaround time.
If Kayak forced users to search for a flight, enter their email, and then receive a list of potential flights minutes later, it would not have reached the level of success it has.
If OpenTable did not allow you to easily filter by party size, time, and location - AND - book in the same session, it would not have reached the level of success it has.
Discrete apps often have lots of competitors and the BATNA or best comparable solution is often so ingrained in a user’s life that they are given a small window in which to prove their value. They are not sought after for their novelty or social status. They are sought after for their utility and ability to solve an infrequent (even if regular) painpoint.
 I originally was going to label these apps as “utility apps,” but realized that was misleading. There are “utility” apps that people use frequently and continuously… you may have heard of a popular one called Spotify.
 Default Dead or Default Alive? Good essay.
 In near direct proportionality to frequency of use.