Most software products need an interface. That interface can come in different forms, but usually boils down to either an installed program, or a browser-based web application. For the desktop (mobile is another issue entirely), web apps seemed to have the upper hand, but successful newcomers — like Slack — and old timers — like Skype — indicate that the issue is still unresolved.
As the founders of Front, an app helping companies manage grouped email addresses (think support@company.com, sales@company.com, etc.), we maintain both desktop apps for OS X, Windows, and a web application. People often ask us the reasoning behind this decision, so here is our answer.

Why would you even consider not building a web app?



The Web has a lot going for it:
  • You code once, it works everywhere
  • There is nothing to install, nothing to update (very handy if your users need an authorization to install software on their computers)
  • The talent pool for web stacks is bigger
  • It makes your product look modern
  • If it works for Gmail/Salesforce/Zendesk, then it should work for you, right?
Those arguments are far from minor, and heavily weigh in for web apps. But for some products, taking a less popular path might be worth it.
Standalone, desktop applications have benefits that are almost impossible to replicate within a browser:
1. Once they find their place in the Windows Start menu or the Mac OS Dock, they are always visible. In the mobile world, an app that isn’t on the home screen is easily forgotten and eventually never opened anymore.If you want your app to become a daily habit in the life of your users, then you cannot take the risk of them not opening it because they don’t see it. A web app that isn’t pinned to the browser is easily closed and forever forgotten. A desktop app has the power to be always visible, and even always open. Don’t miss out on this!
2. Desktop apps are “alt-tab accessible”. Alt-tab is probably the most used keyboard shortcut in the entire desktop universe. For every time someone with your app open hits alt-tab, you get a free impression of your logo and brand name on their screen! How is that for cheap marketing? In all likelihood, your logo will sit between a very popular browser and a cool music streaming service, so the brand association is not bad either. The goal is always the same: your app must become part of the daily routine of your user. Even if they open it by accident, as they’re trying to skip a track in their playlist, it’s still one more chance for you to convince them that you have what they need. Take that chance.
3. also, desktop apps can support download and preview much better than web apps can. Another small but critical thing is the ability to copy things to the clipboard.
4. Another great capability of desktop apps is their easy access to the notification system. Few things get more attention than the bouncing motion of a notified app on the Mac OSX dock.
You can stop checking: it’s just a gif ;)
The bright red circle in the top right corner is so much more effective than ten thousand e-mail notifications combined. It subtly suggests that something interesting happened in your app, and they’re one alt-tab away from finding out what. E-mail notifications get in the way of important communications and once they pass the maximal irritation threshold, are muted forever or marked as spam. Way to leave a good impression, right?
Sometimes those gains don’t offset the pains of developing a standalone app, but it some cases they’re totally worth the effort.
At Front for example, people using the desktop app spend on average 34% more time on the app that those using the web version.

So how do you know if your software can take advantage of those seemingly minor features?



Based on our own experience developing Front, as well as a survey of many software companies, we came up with the following formula. Every software product is different, however when the usage of your software is eitherurgent to your users or preponderant in their activity, then the development of a standalone app is justified.
Here is a my analysis, with a diagram:


By urgent usage, I mean that your users need to react fast to things happening, either inside your app — for instance, an incoming chat message — or outside your app, like when you need to quickly mute a Spotify track to hear the people talking to you. In both cases, you get a significant benefit from having an easily accessible app, that notifies you when an action is required.
By frequent activity, I mean that your users are going to spend a sizable amount of time in your app every day. Anything that helps them do their regular work falls within this category. For programmers, it’s going to be IDEs. For designers, it’s going to be Photoshop & Sketch. Transversal tools like e-mail clients or Evernote also come to mind: you probably don’t use them for the most part of your work, but you still need to open them several times a day and spend some time typing in it. You want them to be always open and not have to sift through twelve tabs to get to them.
If your product cannot really be used in an urgent or preponderant way, don’t burden yourself with a desktop app. The advantages of web apps cannot be overstated, and many great software companies are better off offering web interfaces. APIs like Stripe’s, analytics like Kissmetrics’, infrastructure like Heroku’s, all provide enormous value, but don’t justify the development of desktop applications.
You should know by now whether your software should come as a standalone application or not. It’s time for me to reveal the best argument of the debate: if you choose to build a desktop app, it will come with close to zero additional cost! Recent technologies allow you to build an app as you would for a browser, but wrap it in a desktop app and deliver the same experience everywhere. And there is no dreaded back button, no “open in new tab”, and so on! Granted, it will function LIKE a web app — but it will NOT be a web app, and that makes all the difference. What’s not to like? And if you want to know know how we did it at Front, we’ve shared our experience here.
PS: maybe you’re a bit surprised not to see Google Apps mentioned: after all, some of them (Gmail, Calendar, Hangouts, etc.) induce either an urgent or preponderant usage. However, Google’s long-term interests strictly forbid the development of desktop applications. Indeed, the one and only thing that could threaten Google’s power would be a decline in browser usage: as long as you browse the web, you keep coming back to google.com to navigate efficiently. This is why Google wants you to spend time in a browser. This is why they offer Gmail for free, Chrome for free, Chromebooks at a loss, or why they fund their own competition! So even if users would benefit from desktop Google Apps, don’t expect Google to start distributing them ;