Blog Post

4 tips for developing applications that end users will actually end up using

Stay on Top of Enterprise Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!

People have been espousing the consumerization of information technology for more than three years. But enterprises are still pushing out stodgy apps, and employees are throwing those apps out the window. Fortunately, IT administrators can escape this cycle and give employees applications of real value, if and only if they take steps to listen to end users, Sanofi (s sny) executive Brian Katz explained Monday at the Consumerization of IT in the Enterprise Conference & Expo in San Francisco.

Katz, head of mobile engineering at the drug company, talked about the widespread existence of “crapplications” — you know, those that are technically capable of doing way too many things, or suffer from convoluted interfaces, or take too long to click through — and presented tips for admins to keep in mind in the quest to actually help employees.

  • Companies should find out what employees actually want and need from an application before blindly building one and rolling it out to everyone. “I advocate going on a ride-along, … to see what they’re actually doing,” Katz said. “Spend time with them. Bring them in for focus groups.” For example, his company’s representatives need applications that keep working inside hospitals, where service can drop off. If apps can’t work with that, reps might well delete the apps.
  • Instead of sharing proprietary internal data willy-nilly, expose it through secure application programming interfaces (APIs). And don’t let every user see every file. “We’re going to wrap identity around it, because I shouldn’t be seeing the CEO’s data,” Katz said. Look to API-management tools such as Mashery and Layer 7 Technologies to “appify” your data,” he said.
  • Companies should avoid choosing to write an HTML5 application, a native mobile application or a hybrid based on what they think is hip. “This is a religious war,” Katz said. Companies should do what’s best for their users. An HTML5 application in a browser might not be the best path if a camera function is critical but it takes many seconds for the camera to pop up.
  • Having a mobile-first strategy is important, but it shouldn’t be a mobile-only strategy. It’s a good idea to give employees the ability to work with company data on a tablet and on a desktop. And kicking off a mobile-first strategy shouldn’t be done just to keep to be hip. It should be done to support traditional business goals. That will help guide application development.

So what’s a good mobile app look like in Katz’s book? One, from the New York Waterway, presents users with ferry schedules, maps, alerts, a bus finder, a means to buy tickets and not much more than that. It’s stripped down to what customers want most.

He also gave a shout-out to Shazam, that music-recognizing application. There’s just “a big button that says, ‘Click to listen,'” Katz said. “That’s it. There’s nothing else there. You don’t have to figure out anything. You click it, and it runs. It’s pretty simple. You have a UX (user experience) that people can’t follow? They’re going to get rid of it.”

Just as Box CEO Aaron Levie talks about serving up “a consumer-grade experience” in trying to become the Dropbox of the enterprise, Katz comes across as the kind of guy whs believes that anyone — a consumer if nothing else — should be able to figure out an application aimed at enterprise use. If that’s the case, productivity can increase, and less data will end up floating from internally sanctioned applications to unapproved clouds commonly talked about when people talk about shadow IT.

5 Responses to “4 tips for developing applications that end users will actually end up using”

  1. Another one to qadd to the list is “Listen To The Users And Not The Developers When It Comes To The User INterface”. In my specific accounting field the leader in the sofwtare that caters to this industry made a big leap severa lyears ago from a traditional client/server model to a web based one and instead of trying to take what was working it make it work via the web model they re-designed the thing as if it were a mobile app and of course most users hated it. It took years before they finally got it right and even now, almost 8 years later many users who worked on the client/server system still prefer it because it was faster and more user friendly.

    • That’s because “”client/server” are typically faster and more user friendly. The web was not designed for apps. Even with AJAX and JSON and all the JS libraries – it is still a hack.

  2. Gregory Schwarp

    I work for a large South African bank and my question is, how do you write such applications in heavily regulated industries with complex workflows,p where data quality is critical?

  3. Interesting, but I think startups like Appeos are going to make this obsolete, as they allow business users to create their own applications. This is important, as the users know they want, but don’t want to spend ages writing documents to explain it all to developers, then wait months and end up with an application that was nothing like they expected.

    Business people can build the applications they need, when they need them, and they feel involved, so buy-in is much stronger. It turns out that they can build what they want quicker than they can document requirements in the traditional way. This yields massive savings of both time and money, no more expensive development projects and vastly improved competitiveness, through much quicker time to market.

    It is Enterprise Applications for for 21st Century.

    • This is only partially true. Businesses have been able to build apps for years – with Access (and such) and Excel and other tools. The problem is that these are simple tools and cannot solve complex problems nor enterprise problems (i.e. messaging). While business should be able to develop apps, they should not leave developers out of the loop – otherwise they will be creating another mess for someone to clean up at some point.