After last week's column Is Salesforce.com Wrong to Love Ruby? was published on Forbes.com, I reached out to Byron Sebastian to get his reaction. Byron is CEO of Heroku, which was recently purchased by Salesforce.com.
In the article, I speculated about the reasons Salesforce.com purchased the Heroku platform for Ruby development. I also quoted Byron's statement as CEO of SourceLabs that the enterprise software business model is dead. I thought it fair for Byron to be able to respond to something he said in a different context about 6 years ago.
Ariel Kelman, VP of Product Marketing for Salesforce.com, also joined the call, which turned into a philosophical discussion of Salesforce.com's strategy for building on its rock solid market position.
The interview with Byron and Ariel was fantastic, covering the points I made about the kinds of applications Ruby developers were likely to build and my observations about the relatively small size of the Salesforce.com data model. As usually happens when you get access to the core brain trust of a technology company, a new story emerged, a meta-explanation, if you will, that had two effects:
- First, I am able to explain the acquisition more completely than I've seen anywhere else.
- Second, I understand where Salesforce.com is going and how it is different from the other large software companies aimed at the business market.
In this article, I explain what Byron has learned since he made his provocative statement about the enterprise software model along with his perspective on why Heroku was attractive to Salesforce.com.
In the next article Direct from Salesforce.com: Why We Bought Heroku, I retell the story of the Heroku acquisition based on what Ariel told me about Salesforce.com's master plan for world domination.
The SourceLabs Era Argument: Open Source Will Kill Enterprise Software
In 2004, Byron Sebastian founded SourceLabs, a company that aimed to commercialize open source by offering supported distributions of an open source software stack for use by corporate developers.
I met Byron during this period when I was researching my book Open Source for the Enterprise. He offered the following analysis: organizations could successfully use open source as long as they had a well-defined strategy for closing the productization gap, defined as the difference between the support, documentation, and availability of consulting resources for open source compared with commercial alternatives.
Companies that failed in their use of open source often did so because they ignored the productization gap and had no clear strategy for filling it.
SourceLabs sought to fill the productization gap by offering a subscription service for supporting a well tested stack of open source components that could be used for development. Companies could then deploy open source more easily because SourceLabs took on the heavy lifting of ensuring that the stack was tested, compatible, and supported. SourceLabs would also be a channel back to the open source communities for bugs and feature requests.
Now that I am back in touch with Byron, we will revisit lessons learned from SourceLabs another day. But I recalled when I started thinking about Heroku and Salesforce.com that Byron was surprised that he was able to declare in his speeches that â€œthe enterprise software business model is dead.
What Byron meant by this is that open source had changed the dynamics of how software was created. Communities were innovating in ways that produced products that were often superior to commercial alternatives. The software was available without a license fee.
The cost of using the software was bound up in the cost of closing the productization gap. SourceLabs and SpikeSource took a similar approach to closing this gap, but dozens of other companies, most prominently Red Hat, have shown that the commercialization of open source can be a great business.
SourceLabs never wound up finding enough revenue. SpikeSource changed its direction from its original plan.
What Byron Learned from Salesforce.com
So, was Byron wrong? Is the enterprise software business model dead? SAP, Oracle, and Microsoft are all still around, but they are all pursuing different strategies now than they were then. Every single one of them has embraced open source in one way or another, but more fundamental changes have also occurred.
Byron doesn't think that he was wrong about open source changing the world of enterprise software, but he does confess that it was not the whole story. The threats to the traditional on-premise, license, and maintenance enterprise software were not only coming from open source.
The two factors excluded from Byron's analysis were the change in the software consumption model brought about by software as a service and the dramatic expansion of easier-to-use consumer software.
Byron had a lot to say about these three factors, but here are his comments in brief.
Open source Has Changed Enterprise Software
Open source has changed the way enterprise software works by showing:
- The benefit of open innovation based on freely available source code
- The power of communities to crowdsource various types of development activities
- The value of the freemium model for allowing consumers to try before buying
- The way that vertical integration is not required; the company or community that creates the software does not have to be the same company that provides support
Software as a service represents a new way of consuming software, one that has changed the relationship between the buyers of software, the IT department, and the vendor. Salesforce.com goes to market with a purposefully disingenuous motto: "No Software."
This does not mean that there isn't any software involved; it means that the problems associated with software disappear because Salesforce.com takes responsibility for provisioning infrastructure and delivering software to the end user. The IT department no longer needs to set the stage for supplying software by setting up servers, installing supporting software like operating systems and databases, and connecting everything to a network.
Consumerization can mean many things. In the largest sense, it means that consumer technology has pulled ahead of business software in leading the way forward with innovation.
But it also means that expectations for ease of use and self-service have risen. In the past, end-users may have grumbled about the solutions IT supplied, but they had little choice. Now, with excellent consumer technology at hand that works on any form factor and is accessible via mobile devices, end-users know that software and other technology doesn't have to be complex and troublesome.
End-users are also voting with their mice and using SaaS-based software meant for consumer purposes to support them at work. In essence, consumerization represents the rise in self-service, which makes the adoption and use of technology far more democratic than it has been.
Byron's view is the enterprise software business model is dead but not only because of open source but also because of SaaS and consumerization. What Byron admits he missed was that the changes in the consumption model were far more important than the changes brought about by open source.
If you take a look at SAP, Oracle, and Microsoft, they are aggressively adding cloud-based models and doing their best to make their software easier to use and accessible via mobile networks.
Where Byron may have been wrong is in the pace of change of business models. The enterprise software business model may be threatened, but it isn't going to disappear very quickly.
Why Heroku Thinks Salesforce.com Bought It
Byron's take on why Salesforce.com wanted to add Ruby to its platform fits in with the Cloud2 story that Salesforce.com is promoting. Cloud2 adds social networking and mobile capabilities to the infrastructure-, platform-, and software-as-a-service capabilities of Cloud1.
Just as the adoption of the Java language took place as the Internet became the platform for development, so Ruby is rising as the language associated with social networking and mobility.
By adding Ruby to Force.com and Database.com, Ruby developers are able to bring their talents into a context and a marketplace full of millions of people and hundreds of thousands of companies.
The question I asked in my article was whether Ruby developers will use the Salesforce.com platform. My instinct is no, not right away, and the applications that they build will not be helpful to the business community. Byron doesn't have a strong argument that Ruby developers will suddenly focus on business needs.
But, to be fair, Byron is not the expert on why Salesforce.com purchased his company. Salesforce.com is. In the next part of this discussion, I report on what Ariel Kelman had to say about why Salesforce.com purchased Heroku. The discussion reveals a much more sophisticated analysis of the role of Ruby.