Enterprise app stores need to grow up

Profile picture for user gonzodaddy By Den Howlett September 5, 2013
The need for enterprise add ons is self evident. But the enterprise app stores seem far from chock full of exciting solutions. What's the problem? How can it be solved?

Vinnie Mirchandani has long held the view that Apple is the gold standard for app stores. He believes enterprise tech vendors should actively encourage the building of thriving developer ecosystems. I agree. At least in principle.

For the princely sum of $99, anyone who has what they believe to be a bright idea can submit an application to the Apple AppStore, and then subject to Apple's ever changing approvals process, launch into the big wide world. The model has proven so popular that back in May, Charles Arthur reported:

Nobody knows which was the first app to be downloaded from Apple's iPhone App Store on 11 July 2008, but the total number of downloads passed 50bn on Wednesday, signifying the rapid growth of a business created by the explosive spread of smartphones over the past five years.

Bringing the story up to date, Tristan Louis asserts:

Apple, at its WorldWide Developer Conference, talked about 1.25 million apps in the app store accounting for 50 billion downloads and $5 billion paid off to developers in the last year. To the company, it is a sign of pride to be able to pay this developer community.

Google, with its Android platform has also done remarkably well:

At Google I/O, the largest Android developer conference, Google touted 150,000 developers responsible for over 800,000 apps. While the company does not break out revenue numbers on their apps, recent data in their financial filings seemed to indicate somewhere around $900 million in pay-outs to developers “over the last 12 months” and discussions with external research analysts put the number of downloaded apps from the Google Play store at around 48 billion, close to what Apple has claimed.

Not to be left totally out of the game. Microsoft joined the bandwagon, recently claiming 160,000 apps in its store from 45,000 developers.

However you cut these various numbers and discount the trivial applications, it represents a cornucopia of end user consumer choice. But what of the enterprise world?

Salesforce.com has the most visible app store of any large vendor. According to the latest data (see image below), it claims 1,921 apps and 1.965 million installs. On the side, I am betting that Marc Benioff, CEO Salesforce.com will announce 2 million installs at the upcoming Dreamforce. Add in the claim from Force.com that 100,000 companies are using applications developed on that platform and you might think the enterprise world is also well served. But it isn't.



Despite claiming more than two million developers in its ecosystem, the SAP Store is woefully thin on third party solutions. That cannot be right. In back channels, SAP developer advocates speculate there must be many thousands of custom applications inside the SAP customer base. I have no doubt that is true. The last year, Jon Reed and I have documented 60 plus of the best SAP HANA, cloud and mobile startups out of a total of some 640 plus that have been through the SAP startup program.  If you're a Microsoft Dynamics shop then you are likely well served - at least as a starting point. In the Oracle world, no-one knows how many custom apps exist. It must run millions.

But to Mirchandani's point, the public numbers in the Big Boys tent seem decidedly lame when compared to their consumer cousins. Why?

While the lines between consumer and enterprise continue to blur, most enterprises I come across would like to believe their application landscape is unique and therefore requiring customizations. I don't believe that. Rather, I see plenty of usefulness in configuring applications. The distinction is important because configuring should not require touching the underlying application code where customization implies hard core development at the application code level. That is where cloud applications have an arguably compelling advantage over application code the business can mold to its own purposes. It's also a distinct disadvantage where a business process is genuinely unique, requiring the development of an application that may have no market beyond the unique use case.

Problems, problems

My sense is that there are a set of inbuilt impediments to encouraging developers onto enterprise application platforms.

There is something of a chicken and egg issue. The world is stuffed with great ideas but how to reach the potential market? No market reach = none or limited growth revenue.

  1. The enterprise players have not figured out how to help partners market to the widest possible audience. My feeling is they don't know enough about their customers to segment them adequately to at least start the targeting process without expending significant time and effort.
  2. The enterprise systems integrator world has enjoyed a healthy living from inventing and re-inventing the wheel in the name of meeting unique requirements. Why would they choose to kill the golden goose?
  3. Enterprise buyers are often conservative and will think more than twice before committing to a small developer shop.
  4. Even if a small developer shop has a market busting solution, how would it service potential demand without buckling under the inevitable cash flow pressure?

I am sure there are many more issues but those are my top of mind items. Is there a solution or are we condemned to a rinse and repeat of the last 30 plus years of custom? There are no easy answers.

Any solutions?

The Salesforce.com model points in the right direction and makes clear the potential and popularity of add-ons/extensions. But it is only a start. My sense is that the large developer shops are faced with a problem of truly epic proportions. Here are some examples. How do they...

  1. License in a way that frees developers to do what they need yet continue feeding the funnel of enterprise sales without running into intractable compromises?
  2. Offer developers the opportunity to develop on their favorite platforms rather than de facto closed standards?
  3. Provide developers with the best way to reach markets without crippling cost?
  4. Convince or underwrite small developer shops such that large enterprise buyers can have confidence investing?
  5. Develop a model that removes friction from the onboarding process, often seen as a blocker to success among firms I've spoken with?
  6. Refactor existing platforms and applications to allow developers the broadest possible appeal?
  7. Choose which of the above needs tackling first without destroying their own business models along the way?

It would be great to wave a wand so the problems I identified (and many more) magically vanish. But it ain't going to happen any time soon. In the meantime, I worry that rather than pressure their suppliers, enterprise simply shrugs and either puts up with less than optimal solution landscapes or employs its own developer teams, effectively maintaining the status quo. Neither of those scenarios seems right for the 21st century digital enterprise.

Disclosure: all vendors mentioned other than Microsoft are partners.

Featured image: © Yabresse - Fotolia.com