We use cookies and other similar technologies (Cookies) to enhance your experience and to provide you with relevant content and ads. By using our website, you are agreeing to the use of Cookies. You can change your settings at any time. Cookie Policy.

The Connected Web

Phil Wainewright

2010 Watchlist: People-Oriented Architecture

Vote 1 Vote

With hindsight, I believe we can look back over the decade just finished and see ten years in which a fundamental transformation took place in the culture of computing, one that finally acknowledged the central role that people — the ultimate end users — play in computing. That's a shift that will create important new opportunities over the next few years.

Most of the history of computing has been the story of a quest to eliminate the human element. The very invention of computing was fueled by a need to do repetitive calculations faster than humanly possible. Shared and networked computing evolved largely to eliminate the errors caused when humans manually transferred information from one computer system to another (operations commonly derided as 'swivel chair integration' or 'sneakernet'). But many technologists seemed to believe the science fiction rhetoric of popular culture (from 2001: A Space Odyssey to Terminator and many more) that machines would ultimately make humans redundant. Enterprise computing increasingly sought to exclude all human contact with any data or processes, with the exception of a few highly trained high priests individuals with specialized knowledge and expertise.

The decade of the noughties has witnessed the rise of a countervailing trend, one that prizes human interaction and puts people at the center of computing rather than leaving them at the periphery. Twitter founder Biz Stone summed up this cultural shift in an article this week looking back on 2009:

"For us, it has been a year during which we realised that no matter how sophisticated the algorithms get, no matter how many machines we add to the network, our work is not about the triumph of technology, it is about the triumph of humanity."

This is a trend that I've taken to calling The democratization of IT: overthrowing the complexity and exclusivity of 'Soviet-era software' to put the people that actually use computing, as opposed to those that make or manage it, in the driving seat:
"... the pendulum is starting to swing back towards acknowledging the qualitative role that human beings play in processes and operations. The technology is no longer trying to keep them out of the equation. Instead, its role is to provide automation that helps them do a better job of collaborating with each other, focusing their motivation and acknowledging the experience, insight and creativity they bring to their roles."

I've used the term people-oriented architecture to make a deliberate contrast with our experience of service-oriented architecture in the past decade. Unlike SOA — which too often sought to remake the way that computers talk to one another without any reference to or consideration of user needs and business results — people-oriented architectures have to be developed collaboratively and iteratively with users and business owners, giving them as much freedom and autonomy as possible to control and manage information and processes to achieve the results they want. It's an acknowledgement that people are both the commanding providers and the ultimate end consumers of any of the services in a computing architecture.

Here are some of the growth areas, fueled by people-oriented architectures, to look out for in the coming year:

User experience (UX) design. The hero programmers of the coming decade won't be those who build new operating systems or network protocols. Instead, we'll be celebrating those who build visually inspired, instantly intuitive and effortlessly productive user interfaces.

Iterative, collaborative development. The ability to work remotely while in constant contact and collaboration with users and process owners will become the norm for application development, and any techniques and tools that make the task simpler and more productive will be prized.

Ad hoc, custom computing. People want to be able to grab and adapt automation tools on demand. As a result, we're seeing massive expansion in cloud computing, enterprise mashups and on-demand development and customization.

Enterprise 2.0. All forms of social computing — or 'collaboration', as people in the enterprise sphere prefer to call it — are expanding as people find new ways to productively share knowledge and information either within companies, or between companies and their partners or customers.

Social CRM/HRM/ERP There's been a lot of talk recently about Social CRM (listen to my podcast with Bob Warfield on From Soviet-Era CRM to the Social Fabric of the Web). I think we'll see the same emphasis on real-time information sharing and collaborative feedback expanding into all forms of enterprise application, including ERP, just as we've seen it already begin to creep into various aspects of HR and talent management, from social recruiting to performance management.

Also on the 2010 Watchlist: Ubiquitous Collaboration and Mining the Off-Peak Cloud.


People Oriented Architecture sounds nice Phil but its still the language of a high Priest of computing talking to other high priests of computing. Isn't the democratisation of IT really about the users getting on with it & not worrying about what its called?

Good point, Andy. I guess the reason I've chosen the term here is to appeal to the high priests of computing who read ebizQ, who I'd like to encourage to be more people-oriented in the way they think about computing and how to deliver it.


Fine post Phil!

We have indeed tried to automate-it-all. And horribly failed, fortunately (no Skynet yet)

We are slowly growing from business-rules-based to business-exceptions-based. Especially in the EAI field I've noticed that user intervention is an apparent necessity, as EAI is fully automated and makes the quality of data explicit to all

Funny thing is, I think real SOA (not the SOA we've seen so far) will enable your POA: small new federated services with their own datastore should fill the gap between standardised software and dynamic people/business demands. And they should be flexible like jelly too, not extremely rigid as the current IT-centric SOA services

Just before the holiday break, in a conversation with a colleague I attempted to draw an analog between SOA and the growing confluence of portals, KM, E20, intranets, office productivity apps, etc. (POA?). The commonality I see is that both should ease communication, SOA between machines and POA between people.

Martijn's comment that SOA may yield POA appears aligned with that view as SOA can be how POA is enabled at the system level. Another supporting fact is how the leading E20 platform providers expose RESTful or similar APIs for the express purpose of easing integration into other information silos.

I agree with Phil that we're on the cusp of a change in how IT provides tools to its users. Users, having been exposed to consumer oriented products that increase their personal productivity, will be clamoring for similar tools in the workplace. I think that the answer will require systems, hardware and software, to replace isolated, individual tools directed at specific tasks.


You have identified a critical turning point in the evolution of technology. A correspondent comments about the democratization of technology. Ironically, at the same time that software begins to appear more "democratic", the role of theory is every more important.

Language is democratic, insofar as it is a technology in which we all participate. But to get machines to do language to enormously difficult.

My own interest, as a career B2B sales person, is how awful is most, or even all, sales software. In search of an explanation for this situation, and in dialogue with profs at my local university, I have been steered to the question of ontological modelling.

It is my own view that we are at a place where further progress in building great software -- software the supports the POA of which we all dream -- will be very difficult. Incremental enhancements to individual applications will take us only so far.

The journey to the vast plain of integration is, I would suggest impossible without further progress in modelling, specifically and most importantly the modelling of work. "Work" is of course the subject of almost all business software.

At some point we bump into language -- which would also be the point of real democratization. The Semantic Web is a good step forward here, but only a tiny step.

Here is a short essay I wrote on this topic, almost 10 years ago, after attending FOIS '01: www.personalontologies.com.

For my day-to-day B2B sales efforts, I would dearly love our software boffins to make some real and practical progress!

John Morris
Business Decision Models Inc.
Toronto, Canada

Great post, Phil.

I don't have a problem with the term People-Oriented Architecture.

I have spent the last year researching emerging technologies for my new book, The Next Wave of Technologies. It's obvious to me that the high priests or CXOs can decide to build whatever they want. SOA, agile methods, clouds, SaaS, ISvs, and other concepts allow for so much more flexibility than "Enterprise 1.0."

Ultimately, as you and others point out, end users will drive how new apps and technologies need to work. I truly believe that we're entering a period in which just about anything is possible with computing. The shackles of desktops and laptops, traditional networks, and client-server architecture are being removed.

We are entering an exciting time.

Phil Wainewright blogs about how businesses are using the Web to get better plugged into today's fast-moving, digital economy.

Phil Wainewright

Phil Wainewright specializes in on-demand services View more

Recently Commented On

Recent Webinars


    Monthly Archives