Why aren’t App Constellations working?

Antipattern: when scrolling down the page, at some point it has a popup asking for your email address.

Having read the article, it doesn't mention that unbundling is a cost to the user (more real estate, more task switching). People put up with it for Facebook because it's so widely used already.

If you're not unbundling features but instead launching a new feature as a new app, then you have the same problem of trying to achieve popularity takeoff as everyone else.

Sending data back & forth between web or desktop apps is so much easier than with “mobile apps”. Even on the same device, apps are just too isolated. The platform vendors do everything to disrupt what makes collaboration work elsewhere. They need to destroy the round wheel so they can sell you a trilateral crap version of it for N times the price.

This is actually an interesting question.

In a lot of these examples, "app constellation" means everything the website does is too complicated so it needs to be broken down into several different apps. It's possible that this is just a bad question.

Apps originally designed for mobile don't have problems that you would consider solving with an "app constellation." They just don't have more features than comfortably fit into a mobile app. Tinder vs OKcupid. Twitter/Whatsapp vs Facebook. Skype vs Whatsapp.

I mean, the mobile only (or first) essentially solve the same "problems."

So Facebook needs to solve some hard, ugly problems that will probably never have a tidy resolution while whatsapp doesn't. No wonder they bought them. It's hard to move to less complexity. Just to give one example. FB (or Google+) need to give their users complicated privacy controls to decide who sees what. Twitter or Whatsapp don't, most of it is implicit.

Provably false with the extreme example: Google.

Gmail, Maps, Hangouts, Search and a few others are regularly in the top 100, if not top 20 apps in the iOS app store. Thats at least 5 apps that stay on top. Several games are able to d the same for short durations of time as well... like the derivatives of Angry Birds and a few others.

The constellation works with a series of self established services. It only works as a spin off of the spin off is actually BIGGER than the core app, i.e. FB messenger. Otherwise you should just be looking to kill off the features you were looking to spin off.

As apps grow, and features increase there is always a pressure towards a "mega app” of many features and an opportunity for the "micro app” of focused features. The old adage of “Desktop apps are novels and Mobile apps are proverbs” may not hold as mobile matures, there is nuance.

Mobile applications are especially sensitive to this pressure of feature growth since the users are often interacting in short “mobile minute” spurts of need at a specific time and location ( of course not in all cases, youtube iPad engagement is much longer). The juxtaposition of large apps that support the desired feature and a small app that just does the one desired feature is where larger app stakeholders are trying to hold user base by providing the best of both words by splitting up apps and creating app ecosystems “constellations”.

Some things that are not discussed in this article that discounts the current state of mobile apps is the new integration/federation capabilities provided by mobile platforms such as deep-linking, custom actions and widgets.

Another interesting situation is how a “north star” or “keystone” app may fit into this by providing the highly valued "single sign on” authentication feature for an app ecosystem. I despise login screens and passwords fields in mobile and its Im happy to install a second app so long as you federate authentication.

Maybe in the end one app makes a product but two make a platform ( or maybe just skip the ‘app’ all together and make an API )

It kind of works with google docs since they extracted the editors/viewers for the different file types into different apps. But there is no choice, either you have the spreadsheet app or you can't see the spreadsheet. That doesn't get against the expectations of users trained to use Word, Excel, Powerpoint, etc for different files. However having an app for the front page of HN and a different one to read and write comments would be weird. I guess somebody won't install it.

Define "working". Maintaining #1 in download ranks is a very rough measure and without correlation to something else I wouldn't place a huge value on it. A million people download and then remove it. What then? I'd be more interested in the CLV.

Could I make a case for "app constellations were always a kind of dumb idea"?

For the most part, they're a solution in search of a problem. A designer wanted to make a really clean interface, and in service of making a really clean interface, they killed the genuine synergy of a multi-functional app.

The only case where app constellations really make sense is where the original app was really, genuinely, for-real several apps in one already, with no significant crossover. Even then, unless your customer base are either:

a. Already naturally segmented into people who use functionality X exclusively and those who use functionality Y exclusively.

or

b. Really heavy users of both X and Y such that they'll "pay" the cost of downloading both apps separately.

then it still doesn't make much sense. If you have a function in your main app that people, like, kind of use every now and again, then if you separate it out into a different app, they'll just go without that functionality.