Regular

posted 24 Nov 2006 in Volume 10 Issue 3

The knowledge: Charles Armstrong

Sandra Higgison talks to Charles Armstrong about his study of how information is shared in small communities and the lessons that big business can learn from it.

For most people, the absurdities of the workplace trigger little more than a disbelieving roll of the eyes and a slow shake of the head. If you work in a large organisation it is probably not uncommon to feel as if you’re wading through treacle as you receive mountains of irrelevant and questionable information, try to fathom the latest corporate initiative or despair at the lack of communication.

Most people will simply attribute these frustrations to the foibles of the corporate world – all companies are like that, surely? Charles Armstrong, CEO of Trampoline Systems, however, felt compelled to determine how organisations could distribute information more efficiently.

This curiosity led him to St Agnes, one of the Isles of Scilly off the south-west coast of England, to see how information is shared in a small community. One year later he started to translate his findings into software technology intended to help make corporations share information with the same ease that small villages do.

Armstrong’s experience

Described as the next generation for knowledge management, Armstrong’s experience and flair for information management and ethnography leaves high expectations among those he meets. It was while studying social and political science at Cambridge University that he became interested in the study of human cultures.

“One of the things that attracts me is its holistic approach,” he says. “You’re always looking at a whole system rather than one slice of behaviour; you’re open to significant patterns wherever they might come from.”

Much of Armstrong’s early career was spent in the social sector where he became a fellow of the School of Social Entrepreneurs and met Michael Young, later Lord Young of Dartington, who would become his mentor.

Young can claim the credit for many achievements, including the drafting of the 1945 Labour party manifesto, coining the term ‘meritocracy’, forming the consumer rights group, the Consumers’ Association, and providing the vision for the Open University, a distance-learning institute of higher education established in 1969, which became the model for many other, similar distance-learning institutes worldwide.

“He was an incredible example of how you can take social research and do something practical with it. An academic outsider, he did his research in his own way and rather than let it gather dust he found something useful to do with it. I was lucky to have him as one of my mentors.”

Young is not the only ‘name’ that has influenced Armstrong’s career. In 1995 he worked on a project with David Owen, then the European Union’s peace envoy in the former Yugoslavia. Not only did this work provide an initiation into software development, but it was also the first time he worked with his future business partner Craig McMillan.

“We had to bring together the proceedings from all the conferences, including any maps and documentation connected to the peace process. We even managed to blag an hour of video from the BBC and ITN [Independent Television News], and had support from the United Nations and [computer maker] Apple,” he says. “We were able to publish this thing more or less in real time.”

Given that this project took place during the early days of the web, Armstrong’s description of it as an adventure is apt in many ways. After that, joining a web agency that was then the largest in London, he came into contact with major corporations and government departments for the first time.

“I had naively assumed that information distribution would be more efficient within these types of organisation,” he says.

He couldn’t have been more wrong. “I wanted to know why knowledge was not getting through to the people who it was obvious could do something useful with it while, at the same time, they were being deluged with information that was of questionable worth to them. Most people don’t see the oddity in this and just get used to it.”

A year in the sun

Armstrong observed that small villages are much better at sharing gossip with everyone who’s interested in it than big businesses are at sharing vital information. However, he didn’t understand why. Growing up in Cornwall in south-west England he had enjoyed childhood holidays in the Isles of Scilly, a small archipelago 28 miles off the Cornish coast.

So Armstrong spent a year on St Agnes to study patterns of information flow within the community. “The islands have been going through a difficult economic transition. Through most of the 20th century the main economic activity was flower farming. But that is collapsing under the weight of global competition and tourism is taking over.”

There were several interlocking parts to Armstrong’s work on the island. In addition to his research into information distribution and decision-making processes he also helped set up an initiative to stimulate digital-sector skills in small enterprises to grow alongside tourism.

“This meant that I wasn’t just there as a clipboard-wielding spy; I had a productive role in the community that naturally brought me into contact with people and wasn’t threatening.” His agenda was informal. “I didn’t have a big questionnaire, but allowed myself to be led by my curiosity. When I saw a phenomenon I couldn’t explain I observed it more to see if it occurred again and to see if I could start spotting patterns or anything in the surrounding behaviours to explain what caused it.”

Conducting more formal interviews towards the end of the process, Armstrong formed a number of hypotheses, which he tested to ensure he was on the right track. “I was looking for the smallest number of underlying patterns that would explain all the major phenomenon I was observing.”

Armstrong gives the example of a resident returning from neighbouring island St Mary’s with news that there wouldn’t be a boat on Friday evening, as expected. St Agnes is two kilometres by 1.5 kilometres, has a population of 80 with one pub, one shop, no hotels and no cars.

The news is only relevant to five people, but within hours they knew there would be no boat on Friday. After his year on the island Armstrong had identified the five underlying behaviours behind this kind of dissemination of information:

1. Groups have implicit authorisation parameters that govern how information is relayed;

2. Groups pool intelligence on who needs to know something and who is best placed to pass it on;

3. Groups can function as targets for relaying information;

4. Each person and group is identified with certain semantic triggers that activate relaying behaviour;

5. The ‘further away’ two people or groups are in a social network the higher the threshold for relaying information between them.

Although initially driven by his desire to understand how a community performs these processes, after six months he had started to think about how these patterns and dynamics could improve the way corporate information systems work.

Trampolining

Since completing the project with David Owen, Armstrong and McMillan had worked together a few times. “We regularly discussed what we could do in the space where my ethnographic observations and his understanding of software development met,” he says. “It got to the point where we had created a top-level architecture and realised we could start doing it for real. We raised some grant funding in 2002 and built a simple prototype, which led to the seed financing that helped us set up Trampoline Systems in 2003.”

Based on Armstrong’s findings from St Agnes, the technology is intended to help free organisations from the shackles of illogical and inefficient information dissemination.

By hooking into a company’s knowledge repositories, whether they are e-mail servers, contact or content management systems or databases, the software identifies themes and measures the relationships between information and individuals. Using the 250,000 corporate e-mails that were released to the public following the Enron scandal, Armstrong demonstrates how the software works on enterprise data.

By looking at the e-mails of Enron chairman and CEO Kenneth Lay, the software, called Sonar, shows his main correspondents and most discussed subjects. Sonar combines social-network analysis and semantic analysis to display the network visually. If you were interested in the public report, a strong theme among Lay’s e-mails, the system shows you who he discussed it with, while highlighting any relevant documents, contacts inside and outside the company, and how to reach them.

Each organisation will want to plug Sonar into different repositories and types of data at various levels of detail to reflect their cultural nuances. A company’s information can be politically sensitive so the system can extract social-network data from an e-mail server without doing any semantic analysis.

The system can demonstrate employees’ expertise by bringing together their contact details from one system, show any articles or research they have uploaded to the content management system or intranet, and illustrate their key relationships according to their e-mails. Sonar is also able to identify influence according to forwarded e-mails, quoted documents and other examples of secondary activity.

So a piece of research will carry more weight if it is forwarded by six people who also have a high level of influence.

From St Agnes to San Francisco

Armstrong is grateful to Warren Langley, then president of the Pacific Stock Exchange in San Francisco. By coincidence, Langley happened to be holidaying on St Agnes when Armstrong was doing his study and became interested in his work.

“He highlighted the potential my ideas had for enterprises. When we had a prototype he helped arrange a whistlestop tour of the US where I spoke to about 50 technology entrepreneurs, consultants, analysts and investors. It was a reality check to see whether the idea had merit, how to develop and structure the business, and pace the investment.”

And while Armstrong did not receive a single negative response to the technology, some of his ideas for building the business received constructive, if brutal, criticism. “It was a powerful experience that left me with a mix of intimidation and excitement about the future.”

Since starting the company in 2003 Armstrong has learnt that software development takes time. “It’s not enough to have the right idea,” he says. “The way you execute counts for at least as much if not more.”

The other factor that only became apparent to him last year was the importance of authenticity. “Our response to our early feelings of intimidation was to try presenting ourselves in the most conventional way we could. It was only last year that we realised it was the wrong thing to do and we finally had the confidence to let our personalities shine through, which is when people started to take an interest in us.”

Armstrong cites the intellectual thrill of his work and the feedback the company has received to its work as strong motivators. “In the next 18 months, if things continue as they have done, we’ll be a less intimate business.” The positive reaction to Sonar’s inaugural demonstration at the O’Reilly conference for emerging technology earlier this year confirmed to Armstrong that all the years of effort had not been misguided.

“That audience really knew its onions and verified to me that Sonar will be a huge strategic move for us.” Leaving Armstrong’s loft-style (albeit on the ground floor) office in London’s trendy Shoreditch, I hope that the company’s growth will not cause it to lose its distinctive quirkiness. But with technology that can help solve such obvious problems – the endless productivity drain, duplication of effort and lack of collaboration – the company will undoubtedly grow. Indeed, if Armstrong’s aspirations are to be fulfilled, we will be amazed to think that knowledge management was ever done any differently.

Charles Armstrong can be contacted by e-mailing charles@trampolinesystems.com.

Curriculum vitae

Name: Charles Armstrong

Place of birth: Portsmouth, England

Education: St John’s College, Cambridge (Social & Political Sciences); School for Social Entrepreneurs, Bethnal Green

Employment history: Managing producer, The Electric Company, 1993-97; Consultant, Online Magic Ltd, 1997-98; Custodian, CIRCUS foundation, 1998-2003; Founder & CEO, Trampoline Systems: 2003-present

Personal strengths: Seeing the big picture; an eye for detail; insights from different spheres; calm under fire; persistent

Must improve: Propensity to put off administrative tasks; too intolerant of restrictions

Can't live without: Decent coffee; exploring new places and ideas; friends

What I do to relax: Play Bach fugues; bike across London; escape to Sicily; dance to quirky electronica