Like just about everyone else I know, I have a number of New Year’s resolutions. Exercising more, eating better and staying in touch with family and friends are just a few on the list. And like many of you, I hope this year will be the year I can actually say I managed to keep them for 12 months straight. At least I have Facebook to help me with the last one. I also have some work-specific resolutions: doing more peer review for colleagues on their research notes and coming into the Boston office more frequently to help foster in-person collaboration and discussion/debate of new ideas. Although most of my colleagues are remote, there’s still a group of local analysts, and I enjoy talking with them face to face.
One new idea that’s been thrown around quite a bit lately (both in person and virtually) is applying the “app-store” concept, familiar to people with smartphones, to the world of enterprise applications. Consumers love this concept because it allows them to find, purchase, deploy, use and maintain applications in an amazingly easy way. And because the consumers are searching and buying, they’re likely to frequently use and buy more applications. The concept has been wildly successful for Apple, and now it’s being extended to people with other brands of smartphones. This success has raised an obvious question from consumers who are also users of enterprise software: Why can’t it be this easy at work?
There’s an App for That
A number of IT shops have had that question posed to them or are raising it themselves. Most organizations have more applications than they know what to do with. Wouldn’t this conceptual model help us organize, deploy and maintain apps more easily? And wouldn’t being able to track downloads and usage help us plan better for the apps in which we should invest more, perhaps even the ones we should be sunsetting? It’s a very compelling idea, one that software vendors and service providers are thinking about to help sell more software and/or services.
However, while I agree that the idea has merit and is worth exploring, I can immediately see at least three hurdles that must be overcome to make this idea work (I’m sure there are more). They stem from the differences between how the consumer app-store model works and the reality of enterprise apps for most organizations:
- Homogeneous model versus heterogeneous reality — Of course, my own consumer experiences of the app-store concept is based on owning an iPhone, and I must say it provides a great app experience. The app I need is easy to find, easy to buy, inexpensive (so I don’t need budget approval from my wife) and I get regular updates with new features. Much of this is possible because one company controls just about every aspect, except for the actual code development. Unfortunately, enterprise software doesn’t work that way — companies have a wide variety of platforms and technologies underpinning their app portfolios. The first hurdle is to try to replicate the app-store usability in a heterogeneous environment without breaking the bank on development and integration costs. It would also require a level of cooperation between vendors that, to say the least, has been difficult to achieve.
- Independent versus integrated apps — I have more apps on my iPhone than I care to count. Each one helps me do something very specific, whether it’s checking the weather or playing a game. But all the apps on my iPhone act independently of one another. I don’t have any reason to integrate Angry Birds with Cut the Rope — unless those birds can help me feed candy to Om Nom (sorry, non-iPhone users). I will, however, want to integrate my order management app with my logistics app, so the app store must accommodate the need for information sharing across discrete applications and maintain it seamlessly for end users. Maintaining integrations between enterprise apps is hard enough when IT controls the release cycles. What happens when end users want to upgrade their apps at their own pace?
- Deployment variety — This is somewhat related to the first hurdle, but it has more to do with how to aggregate and deploy the content in an enterprise app store. The mobile app-store model has a very controlled deployment model, where the provisioning, billing and maintenance are handled by one vendor. Poll most organizations today, and they’re likely to have a mix of apps that are on premises and in the cloud. Those wanting to pursue an enterprise app store must develop a standardized approach for provisioning these apps to end users when the apps may reside in distributed locations and have very different license agreements. Keeping track and ensuring license compliance will be critical to make sure your app vendor doesn’t come knocking at the end of the year with a huge bill (like when kids download stuff within games that cost real money).
The Idea Has Legs
The goals and intentions behind an enterprise app store are good ones: Make software easier to deploy and consume for end users. The reality of making this work is another matter entirely. Many software vendors are using platform as a service (PaaS) as a way to deliver their and their partners’ apps in a consistent and structured way. Service providers are exploring what some call “service marketplaces” to deliver more choices and easier consumption for customers. I expect the vendor and service provider community to drive some interesting innovations in the next 12 months to start clearing some of these hurdles. This is a topic Gartner will be digging into in much more detail this year, so stay tuned.
Is your organization pursuing an app-store strategy, or do you think an enterprise app store is a quixotic journey? What role do you see for the software vendors and service providers?