I’ve been a fan of Dion Hinchcliffe for many years. Hinchcliffe operates at the intersection between social, web-based business strategies and web/service-based application architectures – and is a brilliant visual thinker to boot!
In March 2010, Hinchcliffe delivered a very insightful presentation on Social Business Strategies and open business models at the Social Business Summit. I recently came across another presentation that Dion gave in July 2010 at O’Reilly’s OSCON 2010 on the interesection between service-oriented/web-oriented archtiectures, social business strategies, and cloud-based business models. The presentation can be viewed below:
The slides for the presentation can be downloaded here.
In the presentation, Hinchcliffe discusses the role that new Service and Web-based application architectures, deployed in the cloud, play in enabling radically different business models. As a result, the talk is both more technical than the presentation Dion gave at the Social Business Summit, and more ambitious in scope.
In this blog post, I’m going to summarize the takeaways from Hinchcliffe’s presentation that I found most interesting.
Converging models of Open Computing
Hinchcliffe begins the presentation talking about 3 key convergent models of Open Computing: Cloud Computing, Social Business (or Enterprise 2.0), and Service-Oriented Architecture, as communicated in the slide below:
The pervasive theme here is that of open models of collaboration and interaction, fundamentally enabled by the global reach and ubiquitousness of the Internet.
Change is Pervasive in Business and IT – it’s open and it’s social
The following slide speaks for itself:
Change is pervasive in both business and IT today – it’s open, it’s social, and it permeates every aspect of the business from customer service to product development.
Service-Oriented Architecture and Web-Oriented Architecture
I’m going to show numerous visuals from Hinchcliffe’s presentation, all of which originally appeared in previous Hinchcliffe blog posts. I really can’t say enough about how masterful Dion is in conveying complex topics visually.
Next, Hinchcliffe explores the importance of SOA – or Service Oriented Architecture – as a foundation for enabling this new world of open computing on the Web. He notes that SOA emerged in part as an architectural paradigm for integrating disparate application siloes across the enterprise and between enterprises, as well as for designing software solutions “in the very large” – for designing systems of systems of systems.
Services are the most recent paradigm for how applications expose their functionality, as illustrated by the slide below on the history of software:
And while the original vision for connecting services was promoted under the term SOA, it was really the Web that showed us the “true path” towards service-orientation, which Hinchcliffe refers to as Web-Oriented Architecture (or WOA). And while the underlying principles are very much in accordance with SOA design principles, the implementation specifics of WOA are what make it unique. More on WOA later in this post.
The business case for SOA in the Cloud
Increasing, however, SOA-based implementations are found not behind the walls of the enterprise firewall, but in the Cloud. And the reason is that emerging business models – that massively leverage network effects – are built offering cloud-based services. The visual below shows that Amazon now uses more of its IT infrastructure to provide cloud based services to external applications than to services its customers through its website.
We’ve been focusing on applications and point-solutions – especially in enterprises – for a long time. We’re trying to solve a specific problem. Whereas on the Web and in the Cloud, we’ve long since moved beyond apps, and we think primarily about platforms. That’s where the big value is or the order of magnitude, or perhaps several orders of magnitude, comes from …
The network effect is substantially higher when we build a platform that says, “No, we’re not going to be in a single place or a few points of light on the Web. We’re going to be on 50,000 – or in Amazon’s case over 300,000 – different other places. Our supply chain, our Cloud, will be integrated into everything, and we get value out of every piece of that flow.” This is the new economic picture for Software, for Services, for the Internet.
Service-Oriented Architectures enable new Business Models
So the strategic goal of SOA, at the highest level, was turning applications into platforms. That’s what we learned 5 years ago with Web 2.0: A platform beats an application every time.
… It’s a new digital business model. And this is the hardest part we see with traditional organizations that look at there business in a very 20th-century way. It’s not paving the cow path. It’s not rewriting the brochure onto your website. It’s about creating fundamentally new businesses based on the capabilities that you have – what data and what services you have.
For more on SOA (or Services) as a platform for new business models, please see:
Clouds as a platform for new Business Models
Then Dion gets to the absolute heart of the matter:
So one of the questions I like to post is “How do we best think about our Clouds?” Is it about the Services and the Data? Or is it about the Business Models around it?
This question is illustrated in the slide below:
If you look at a World Bank, Best Buy – these are very traditional organizations. They now have full-blown developer networks in scale. Many other traditional organizations do too. And this is one way of really thinking about this, is we see the developer network wrapped around traditional businesses. It looks strange to us, even if we’re inside the industry. But this is really what we’re seeing. This is how we turn our businesses into platforms. This is the new Cloud.
On the value businesses derive from providing Cloud-based services, Dion comments:
It’s not from the first use of that Cloud API [that the value comes], it’s from the hundred or thousand reuses from all the partners you may have. Whether that’s a horizontal cloud service like Storage, or a vertical service like Credit Card Processing. Your job is to drive unintended uses, and make it as easily consumable as possible.
Of course, that’s what we wanted in SOA. But it’s actually what we achieved on the Web. The Web has been a runaway success. Every major Web product has an API and a developer network. This is how you distribute in the Cloud today.
But the focus was different. [SOA] was much more about creating Services, and the other one [Web 2.0 Services] was about driving consumption – how do we make sure what we have in that service is the best, the most valuable. Because in an era where the world is flat, people can vote with their feet, with a URL and plug-in a new API.
So what happened to SOA?
So what happened to SOA? Hinchcliffe’s view is that we are actually seeing very low rates of adoption amongst enterprises – to the tune of low-high single digits. This is due to several reasons, including:
- Failing to drive the business conversation – SOA was good at connecting stuff together, but it didn’t really drive conversation around re-architecting the way we do business digitally
- The technical approach seemed to have the wrong focus, and Web 2.0 organizations found better architectural approaches (i.e. Web-Oriented Architecture)
The contrasting approach of SOA and WOA to application architecture is illustrated in the visual below:
I’m not going to elaborate on the differences between SOA and WOA in this post, but Hinchcliffe does add that most Websites and Cloud Computing platforms today use designed around Web-Oriented Architectures (including using REST-based APIs).
Here’s another visual from Hinchcliffe’s presentation that illustrates key principles/components of a Web-Oriented Architecture:
That’s really what we’ve learned, that Services and Data work much better when they align naturally with the grain of the web. … All of our data, and the entire data model, is nothing but links and pages if you will, Linked Data of course.
For additional insight into Web-Oriented Architecture and its relationship to SOA, please see the following Hinchcliffe articles:
- The SOA with reach: Web-Oriented Architecture – April 2006
- What Is WOA? It’s The Future of Service-Oriented Architecture (SOA) – February 2008
- 12 Things You Should Know About REST and WOA – April 2008
- The SOA world begins considering Web-Oriented Architecture (WOA) in earnest – September 2008
- Unboxing Web-Oriented Architecture: The 6 Aspects Of An Emergent Architectural Style – June 2009
- Where Is The Future of SOA Headed? Where The Web Goes… – September 2009
- A Web-Oriented Architecture (WOA) Un-Manifesto – December 2009
The Cloud IS our Global SOA
So what we’ve learned, says Hinchcliffe, is that the Cloud IS our global SOA. There’s a full spectrum of what we’d call Services – anything that’s pullable from HTTP is a cloud service, like it or not. Even a web page is a cloud service.
But that unstructured information and HTML isn’t necessarily where the interesting part is – although entire businesses have been formed around making sense of that, like Google.
It’s the Services head that is particularly interesting to us – where the Web APIs are. … And we’re now seeing brand new emerging models for distributing services, and I’ll talk about that as we finish. … But we’re now understanding that everything that we can pull via HTTP are part of the Cloud.
Enterprise Data in the Cloud
So all these ideas are allowing us to create our own Clouds.
Quoting Hinchcliffe again from the presentation:
Open APIs and Web-Oriented Architecture have informed the best approach to SOA, that battle seems to have finally been won. And Cloud Computing, that’s standard. Everything I just said, that’s generally how we do Cloud Computing.
But there’s a new scenario – and that’s where we start to talk about the other threads in our conversation – that is creeping into our organization, and that’s data explosion. … And given that the visual web is soon going to be eclipsed by the non-visual, Services “cloudy” Web – if it hasn’t already for most of us, in fact – we’re going to have this same issue, because we’re creating this whole microcosym within our organizations in the Enterprise Data Cloud. And this is really the challenge of what we call Enterprise 2.0.
For more on enterprise data in the cloud, please see the following Hinchcliffe article from February 2010: The Enterprise Data Cloud: Why Information Power Is The Future of Business.
Next, Hinchcliffe moves onto the topic of Enterprise 2.0 – or social media and collaboration within the Enterprise based on Web 2.0 technologies.
However, Dion provides a very interesting technical definition of Enterprise 2.0 in the slide above: The application of user generated Web technologies to create a WOA data ecosystem. This definition very much builds on the Enterprise Data Cloud notion that Hichcliffe discusses in the previous slide. He goes onto elaborate:
Open communication, Shared knowledge – making sure anyone can participate – that’s very important, fundamentally web-oriented, and emergent and free-form. Those last two pieces are very, very important.
On the pace of technological change and its impact on the enterprise, Hinchcliffe has this to say:
Right now technology is accelerating away from the rate of change in the business world. There’s this growing gap, and I don’t know what’s going to happen. This is what all the books are [talking about these days]: What are we going to do about this growing gap? Centralized, bureaucratic organizations can’t keep up with the pace of innovation out there. What’s going to happen? Is there going to be some dislocation or disruption? We don’t know what’s going to happen, but it’s going to be interesting to watch for sure.
The Social Web
The Social Web now is the dominant model for communication, except maybe text messages. … The number of global users of social communication has eclipsed e-mail. The key point here is that e-mail is not based on the architecture of the World Wide Web, but most of the Social Web is. So you look at communications going fundamentally web-oriented – in terms of link structure and how the data’s connected, and how we locate it and parse through it, all the APIs that access it; all that is in now a format that aligns with where the cloud is going, and is what we’re going to be using in terms of communication in the future. This is just the beginning of the trend, and most organizations aren’t ready for this.
The future of Enterprise 2.0
Where we’re headed, according to Hinchcliffe, is towards what he terms a peer-produced Intranet – this healthy web-based ecosystem. Where we can use the applications of our own choice, to create the data and applications we need, and be competitive and connected to the greater cloud as well. And do this alongside traditional enterprise systems that are running the business today.
The Big Picture – Social Business
Hinchcliffe does a quick call-out to the emerging category of Social Business.
Social Business is not the topic of this talk, so Hinchcliffe only lightly touches on the topic. For a full treatment of the topic, please see Hinchcliffe’s presentation at the Social Business Summit 2010.
The IT Department’s view of Enterprise 2.0
So how do IT departments have to adapt to an enterprise 2.0 world? Says Hichchliffe:
Your IT department and your organization is going to look at things much more monolithically. They’re not ready to have this highly web-oriented, every piece of data has an address, every service has a URL, every endpoint is all based on these technologies, and it’s a relatively simple and flat structure where everything is available. I can even find data via my search engine, I can find data via my search engine, if I know where to look for them.
Unfortunately our enterprises are far too complex for that. This is all the typical high-level moving parts in a modern enterprise around data.
You have the archiving, and the auditing, and the search, and the visualization, and the business intelligence … And we’ve got to somehow break away from that. We need governance and security, but the world we’re moving into right now is going to be much, much more federated. The applications of the web already do that today, and those same models are moving into the enterprise.
For a very nice analysis of the importance of data-centricity for today’s businesses, please see Eight Reasons Why Data-Centricity Is The Future Of Business.
The Enterprise is not the Web
But we’re going to have to help our organizations understand that. That’s going to be one of the biggest roles of education, because the enterprise isn’t the web.
The following graphic illustrates some of the key difference of Web-oriented architectures applied to the Web vs. the Enterprise:
We want to replicate the successful aspects of the large cloud – that’s the dominant IT system in the world today. But we have to figure out how to adapt and how to augment our systems.
Lack of visibility into Enterprise Data – a key challenge
But the real issue is going to be this – and we see this issue talking to Jive software. They find that customers that have had these types of tools deployed throughout the organization for a couple of years are encountering this exact progression we see here, where you have this information exposure. Suddenly what everyone is doing, and all the data people have and all the projects and documents, and all the information and apps that they’re using, are suddenly visibile within the organization.
And they’re like “Wow, how am I going to deal with this?” And it’s like Clay Shirky said, it’s not information overload that’s the problem, we want the information. It’s filter failure.
Most of the vast repositories of data in enterprises – something like 80%, 90% in some organizations like government – is just not accessible in any practicle manner. Yet we pay a fortune to keep it around, and use it once in certain applications as opposed to making real use of it – the way we do on the Web – which is to get it out there to everyone who needs it.
Enterprise 2.0 is changing that.
Web and web-oriented based cloud infrastructure will create this visable surface area. And we have this explosion I have there on the diagram. The next phase is going to be “OK, how do we deal with it.” All the silos are actually being torn down. Finally. These tools naturally do that – they link everything together, they connect everything together.
All your services can be easily connected together. A little bit of glue in your favorite scripting language can pull together systems around the world – I do demos all the time for web-oriented APIs build amazingly powerful applications using Rails in a few minutes, using APIs from all over the place. That’s just how the world works now.
But that visible surface area causes a huge management issue. Who’s going to adminster that? That’s what I hear- the IT department says “Who’s going to support that?” We don’t have the answer to that, but the answers probably not how we do things today.
For more on enterprise approaches to Open Data, please see The future of enterprise data in a radically open and Web-based world.
Enterprise Mashups – an emergent trend
One emergent solution is Mashups. We’ve been talking about that for 5-6 years. The Web’s been doing this for a long time. But they’re just now starting to get into businesses in a real way.
End-user Empowerment – DIY Computing in the Enterprise
But it’s really about saying, well, in any organization there’s maybe about a few hundred developers, and there’s maybe a few thousand IT people, and maybe 5, 10, or 100 thousand of everyone else. … Where’s the productive capacity? Where are the ideas coming from? Well they’re going to come from that whole spectrum. So that’s where the idea of participation comes in. Authorship needs to be in everyone’s hands.
… These very free-form and emergent social platforms are natural integration points. Cut-and-paste integration goes away. You’re actually using these social tools to knit together things. … Putting authorship ability in everyone’s hands to keep the cloud tidy and neat and organized really works, because it works in a much bigger example we all use every day.
… Civilization advances when things that were hard are made simple. And this is what Cloud technologies are doing, they’re making things that were formerly very hard much simpler.
Enterprise App stores
So the Mashup story is very exciting – we’re seeing it go into the enterprise now. But what’s really exciting – the latest trend, to wrap this session up – is App Stores.
Hinchcliffe goes on to briefly discuss App Stores in the enterprise, but I’m going to wrap up the post here.
In Summary …
A wonderful presentation from Dion Hinchcliffe that ties together disparate threads of Service/Web-Oriented Architecture, Social technologies in the Enterprise (aka Web 2.0), and Cloud Computing.
But I find it interesting to contemplate how leveraging network effects – across both content and ad networks – can be applied to a Community Journalism strategy.
Anyway, that’s that. :)
For those interested in application architecture – and more particularly, Web Application Architecture – you may want to check out some of the wonderful articles published by Dion Hinchcliffe.
The article Tips for Building Next Generation Web 2.0 Applications provides a nice representation of the emerging Web Application stack that includes both outsourced services and social software capabilities, shown below:
Some other great recent software architecture articles by Hinchcliffe, see:
- The SOA with reach: Web-Oriented Architecture
- What Is WOA? It’s The Future of Service-Oriented Architecture (SOA)
- 10 Must-Know Topics For Software Architects In 2009
- Building Modern Web Apps? Better Have A Deep Competency in Web 2.0, Open APIs, Widgets, Social Apps, and Much More
Dion Hinchcliffe delivered a great presentation at QCON in May 2009 on the Web as a Platform for developing application and delivering business services.
The above video presentation displays accompanying slides, but the slides are not available for download. However, the slides for a similar talk given by Hinchcliffe at Web Expo 2.0 in April 2009 is shown below:
As usual, fantastic insight from Hinchcliffe.