STOIC is a platform, and platforms are notoriously difficult to bring to market. This is something that we’ve known intuitively for a long time, but we’re starting to better understand why. The reason was best summarized by someone we talked to last week during some due diligence: unlike applications, platforms deliver very little business value initially. Their value is delivered over time, in a very incremental fashion. Over a long period of time though, the value increases and is guaranteed for even longer periods of time afterward, because of the fact that successful platforms tend to be sticky. But the time-to-value can be quite long, especially when you consider that platforms can have switching costs (cost of switching from one platform to another), hence their initial business value is actually negative, and can remain so for a while.
Unlike platforms, applications deliver most of their value immediately after they’ve been deployed. This could take a while for complex applications like ERP systems, but simpler applications like office productivity suites will deliver value from day one. That being said, this value tends to decrease over time, because better applications invariably become available on the market, and the perceived value of legacy applications diminishes as a result. And in the world of business applications, for better or for worse, perceived value is almost as important as actual value, at least from the viewpoint of the economic buyer who will sign a large check for buying new applications. Nevertheless, the fact that applications deliver most of their value initially makes them dramatically easier to sell than platforms.
Such is the challenge of bringing platforms to market: platforms are great over time, but you’re always better of selling an application built on top of the platform instead. Of course, if you do so, you might end up over-optimizing your platform for the particular set of use cases supported by the application you build on top of it, thereby restricting the applicability of your platform to this particular application. As a result, you end up with nothing more than an application, instead of a generic platform that could support the development of a wide range of applications.
Another approach consists in developing a very simple application to support your go-to-market strategy, while working with partners who develop other applications on top of your platform. The former will bring in a steady revenue stream, while the latter will ensure that you keep building your platform at the same time. If you do that long enough, you should end up with a successful application and a solid platform underneath. It might take a while to get there, but it’s a powerful way to break the catch-22 cycle that too many platforms get caught into.
Picking the right application can be a challenge though. You want it to be simple enough that it won’t divert too many resources, yet sophisticated enough that it will really take advantage of the platform underneath — otherwise, your platform becomes a liability more than an asset, and you’re much better of developing your trivial application on top of a more established platform.
You also want to pick an application for which there is a market, which I would define as a set of people who have money to spend in order to solve a problem they wake up in the morning knowing they have to solve. And you want this market to be large enough that it could support you and multiple competitors for a while, yet not too large that it’s already captured by established players. In other words, you want an application that can address a fast-growing emerging market, with plenty of room for many different players.
Then, in a perfect world, you’d like to pick an application that can effectively showcase the features of your platform, giving users a taste for what they could do with it if they were to peel a few layers of the onion. But you want this onion peeling process to happen over time, at the user’s own pace. In other words, you want the application to be able to sustain itself as a standalone product initially and for a while, yet be enhanced by its underlying platform. Or, coming back to the time-value analysis introduced earlier, you want the value of your application to increase over time, precisely because it is built on top of a platform, instead of being nothing more than a bespoke application which value ineluctably decreases over time.
Last but not least, you’d like the growth of the market defined by this application to be supported by tectonic changes that are happening in the industry, and are not limited to your application’s market. You want these changes to apply at a systemic level, to be positively guaranteed, and to be long-lasting, because it might take you a while to deliver the right application and to make its underlying platform work at scale.
If you can find such an application, and if you can package it in such a way that your platform initially becomes almost invisible, you’re in good shape. And if you can find a way to bring this application to market without having to spend massive amounts of money in sales and marketing, you should find success along the way, no matter how hard you try not to.
Ever since we’ve started STOIC, we’ve been pondering this issue, one way or another. But we could never find an application that fit all the criteria outlined above. Then, over the past few days, a candidate has been slowly but surely emerging: STOIC Drive. We’re not yet certain that this is the application that we’ll use to go to market in the way that I just presented, but it’s looking more and more promising. Here is why:
First, let me explain what STOIC Drive is. In a nutshell, it’s a private clone of Google Drive. It’s something that you can deploy in your private or hybrid cloud, and will give you most of the features offered by Google Drive. It’s not very fancy, and there is a well-defined market for it: Enterprise File Sync & Share. There are quite a few players in that space already, but no market leader yet, and very few (if any) that are focusing on the segment we’re interested in.
Second, let me explain who STOIC Drive is for. It’s aimed at large companies who have just deployed OpenStack to create their own private or hybrid cloud and are now looking for applications that could be deployed on top of it and could reduce the rogue usage of public cloud applications. By far, the most popular SaaS applications are public drives like Box, Dropbox, Google Drive, or Microsoft SkyDrive. Therefore, it is reasonable to think that early adopters of OpenStack will soon be looking for online drive alternatives that could be deployed on top of their brand-new private cloud infrastructure. These early adopters are the initial target that we have in mind for STOIC Drive.
Third, let me explain what problem STOIC Drive solves: on one hand, business users want the convenience of consumer applications like Dropbox, which is the reason why they’ve been adopting them in droves, and the recent BYOD trend has accelerated this adoption even further; on the other hand, IT managers want to keep control of their infrastructure and enforce the variety of governance policies that justify their very continued employment within large organizations. STOIC Drive helps reconcile both needs, by delivering the convenience of consumer applications with the governance of enterprise systems.
Fourth, let me explain how STOIC Drive works. At the most basic level, STOIC Drive is deployed as an application on top of OpenStack. It takes advantage of OpenStack Swift to store large amounts of file-based content on the customer’s private cloud. But STOIC Drive also happens to be an application built on top of STOIC, with a great user interface for both static and mobile devices. In between, it offers a set of connectors allowing you to mount the most popular public drives (Box, Dropbox, Google Drive, Microsoft SkyDrive, etc.) as virtual folders on the drive. It also provides an API to connect to Amazon S3 so that you can do this cool cloud bursting thing that everybody seems to be so much in love with.
Fifth, let me explain what’s so cool about STOIC Drive. By providing this connection to all public drives, it allows business users to continue using them, while having full backups on the private branch of the drive. And by having the drive built as a STOIC application, it adds a host of features to all public drives, including:
- Advanced permissions (Cf. Authorization engine)
- Indexing & Searching
In other words, this drive actually provides most of the features usually offered by traditional Document Management Systems, but with a consumer-grade user interface, and a hybrid cloud deployment model. And because it’s built on top of OpenStack Swift and ElasticSearch, it will give you petabyte-level scalability, while remaining highly cost-effective.
If we summarize, here is what is getting us excited about this application:
- Emerging market with high potential for growth and no dominant player
- Well-identified economic buyer buyer
- Clearly-understood value proposition
- Large addressable userbase including both business and technical users
- Large number of paying users within accounts
- Small number of required features
- Significant potential for showcasing the underlying platform
For all these reasons, we think we have a winner. We also think that picking the wrong application is actually less damaging than picking no application at all. If this one does not work out, we can always build another one. But not focusing on any application is a sure way to fail as a platform vendor. As a result, we’ve decided to investigate this opportunity as aggressively as we can, and to make a final decision during our upcoming company-wide gathering at the end of the month. In the meantime, Pascal is burning the midnight oil to get a prototype ready soon.