Why cloud providers must look beyond services to APIs

Modern applications are becoming webs of services. To deliver real value, service providers need to offer complete platforms along with the services customers demand

business cloud services flowchart

Back in the early 2000s, I worked as an architect for a San Francisco consulting company, designing and building large scale Internet systems.

One project took me to Lisbon, where a new 3G mobile telephone company wanted to differentiate itself from its rivals. Building on what were then new ideas, we came up with an architecture that would allow applications to be built by hooking together services. The only rule was that those services needed to provide a public set of interfaces.

Roll the clock forward a decade or so, and that early service-oriented architecture is the norm, with collections of services and public APIs at the heart of a new generation of enterprise applications. We can pull together components and services from our own line of business apps and from third-party tools and services, linking them together to form applications that are tailored to the way our businesses work.

To meet that demand, a widening array of services has emerged in the public cloud, but it's an odd sort of marketplace. After all, if it's just a service, then it's just a feature -- and no one buys features when they can build their own. How can cloud services add value to their enterprise users? More important, how do they bill for those services?

That pressure is felt most keenly by cloud storage providers. A short sharp pricing war among Amazon, Google, and Microsoft has left all three battered and bruised. Where it used to be possible to build a business around selling storage, the fact that enterprise giant Microsoft has made unlimited storage a feature of its Office 365 platform has left companies like Box and Dropbox rethinking their enterprise business models.

It's no longer possible for them to be a feature. If a business wants to use massive cloud scale storage, it's there with Google Drive or Microsoft OneDrive, along with basic collaboration tools. What's a cloud storage company to do?

The answer lies in the way both Box and Dropbox have shifted strategies over the last year. Instead of being merely additional storage for a business, they're becoming storage platforms. The approach makes sense, and it mirrors the model that cloud telecoms service Twilio has been using to build its business.

With Twilio, there's no product to pick up and use. Twilio offers a much more interesting attraction: its expertise as a telecom provider wrapped up in a set of RESTful APIs. You can sign up for a Twilio account with a credit card and start adding telecom features to your apps. Perhaps you want to add SMS support to a logistics app, so a delivery truck can automatically text a customer when it turns into their street. Or maybe you want an application that can automatically route key workflow tasks to the right person by an appropriate method, so a purchasing approval can be done with a touch-tone phone.

It could have been easy for Twilio to be another feature, another Skype or a Viber. But instead, by breaking apart telephony into a set of APIs, the result is a platform that lets developers pick and choose the components they want to build into their apps.

Becoming a platform like this isn't easy. It means changing customer relationships and building new billing models. Instead of selling to users, you're selling to developers -- and those services need to be reliable and secure if they're to become part of a new wave of enterprise apps.

Box's APIs have been around a while, and the company has been focusing on delivering managed storage services and workflow tools to industry verticals, with particular interest in finding new ways of presenting information to users. Take its recent acquisition of a medical image viewing company as an example. Using Box's tools, it will be possible to build a secure workflow from a scanner to a doctor's tablet, making key diagnostic information more widely available and easy to embed in a range of different tools.

Dropbox's enterprise platform isn't as developed as Box's, but it's rapidly gaining features. There's a strong focus on collaboration and finding ways of embedding your corporate workflow in familiar desktop tools. You don't need to write apps that mimic familiar forms when those everyday Word templates can be made part of your application and when you can programmatically control how the resulting documents are shared and how they're stored (whether they originate in the Web, from an line of business app, or on a user's desktop or tablet). With programmatic support for key information management standards, protecting user information in the cloud becomes a matter of a few lines of code.

As both companies focus shift from apps to APIs, what were once monolithic apps and services will fragment, so enterprise developers can pick and choose the APIs they want to use. Working with developer-friendly tools and services and with SDKs on GitHub ready for download, both Box and Dropbox are clearly aiming to be among the many platforms that form the foundation of new business applications.

It's not all sunshine and roses. There's still the problem of the lingering fallout from the cloud storage price wars. Both Box and Dropbox will need to find new ways of charging for their new services -- an interesting problem for their business teams. Just how valuable is a secured document or an X-ray image?

Copyright © 2015 IDG Communications, Inc.

How to choose a low-code development platform