Royal Dutch Shell (Netherlands/UK)

Peter Kemper arrived for a long awaited meeting with Jeroen van der Veer, CEO at Royal Dutch Shell. There were three things on Jeroen van der Veer's mind. The first was Second Life and how virtual worlds could be effectively utilised at Shell. The second was the recent implemen- tation of wikis at Shell and how effective they had been. Lastly, he was interested in any fur- ther initiatives that could mobilise knowledge more effectively across Shell, such as their earlier investment in SiteScape and development of vir- tual communities. They had 20 minutes to dis- cuss these matters before Jeroen van der Veer was expected at another meeting.

Peter Kemper, Knowledge Management Portfolio Manager (IT Department), was a keen believer in user-generated content and the sim- plicity of Web 2.0 solutions. He started by sug- gesting that wikis had reduced the problem of information overload with much fewer docu- ments. He estimated that there were probably between 300 and 500m documents within Shell archives. This was too great for even the most determined search engines in the world. 'To overcome this problem', he added

'We followed the Google/Wikipedia approach as much as possible: no traditional big bang imple- mentation but entirely bottom up supported by "smart marketing". We concentrated on keeping it as simple as possible (e.g. transparent login/NO access control). On the content management side we linked to our existing discipline structures (e.g. geol- ogy, production chemistry, wells technology etc.).'

The wikis were also connected to Global Networks or discussion threads within Shell. These had been established in 1997 to assist with community collaboration and community self- regulation. He continued that they had added 'information similarity checking' to the wiki environment. This allowed the system to provide a list of existing documents that were 'similar' in content to the wiki. The emphasis was on reduc-

ing the number of documents in Shell's knowl- edge base and eliminating duplicate documents.1 Prior to the wikis, Shell's main media for

maintaining and sharing knowledge had been e-mails and Powerpoint presentations. He gave the example of the geological structures of the Atlantic Margins held as a series of images and added: 'It would be a nightmare to maintain this kind of information in Powerpoint! No offence meant: Powerpoint is good for presentations but not for creating and maintaining a knowledge base.' Peter maintained that the wikis achieved six times more productivity than their former Powerpoint presentations and allowed content to be linked directly together. He showed that wikis had been used successfully in translating four hundred page contracts into crystal-clear prose that could be acted upon by ordinary employees. There was a groundswell in dialogue and activ- ity across Shell where staff were communicating with creators of wikis especially when they want- ed to go beyond the knowledge found directly in the wiki. They did this directly via e-mail or using the discussion pages with each wiki. Authors were also alerted via a 'watch list' func- tion if anyone amended their entry. Peter Kemper offered a picture of progress with wikis at Shell. He confirmed that there were rough- ly 40,000 registered users compared to the overall employee population of 150,000. He noted that the numbers of active users remained relatively constant even though user numbers had increased. He observed that there were three types of users: content owners who created the content, con- tent editors who improved syntax and the overall presentation of wikis and information consum- ers. He reminded Jeroen van der Veer that 60 per cent of all staff would go into retirement over the next eight years and that wikis provided a useful medium for capturing and sharing this vital organisational knowledge.2 All Shell Open University courses had their content success- fully converted into a wiki. This allowed course

members as well as the wider Shell community to access this knowledge. However, Peter insisted that the most valuable knowledge was that found in the heads of co-workers. Wikis had provided the fastest and most effective medium for employees to tap into their co-worker knowledge.3 This know- ledge included aspects of production engineer- ing, technical handbooks and the non-technical aspects of their roles. He tried to illustrate the level of activity with wikis each day. There were around 100 registered new users each day, around 300 edits of wikis each day and over 5,000 views of wikis each day. This represented substantial activity in both knowledge creation and knowl- edge sharing. There were also over 1,300 users who had created new content in wikis since its inception at Shell. Peter Kemper added: 'We have definitely met the needs of many users in the com- pany, who were looking for efficient and flexible ways of sharing their knowledge, and who wanted to co-write and co-publish information'.3

Peter Kemper viewed the emergence of wikis as an extension of Shell's communities of prac- tice (CoPs). In both, insights were shared and they brought people together with common interests. This was important as many Shell employees shared their expertise by moving between different projects around the world. As many as 80 per cent of Shell's technical profes- sionals were away from home at any given time. Hence communities of practice originally devel- oped more through electronic means rather than regular face-to-face contact. At first, there were 107 different CoPs around the world but the dif- ficulty was that there was little communication within them. In 1999 Shell rationalised the 107 CoPs into three global networks: Surface, Wells and Subsurface. In addition, eight CoPs were formed that were cross-cutting across functional areas. These included benchmarking, competi- tor intelligence, procurement, knowledge man- agement and IT. There were around 20 subject experts in each CoP and one full-time global coordinator (facilitator). Sitescape Forum was the social software adopted underpinning Shell's CoPs thanks to its ease of use and administra-

tion. Most messages and requests for information were responded within a few hours.

Peter Kemper tried to demonstrate the costs and benefits of the Sitescape system. In any one CoP, there were around 80 new postings each day and around 350 logins and files viewed by users. Each posting had at least 60 views and most new entries were responded within 24 hours. There were likely to be multiple responses to each new entry as each community had at least 1,000 members. Given the global nature of Shell's operations, users were encouraged to place entries in their own language and to use the language tool to translate information from one language to another. However, users were warned about the pitfalls of vital knowl- edge being lost in translation and encouraged to check source languages with colleagues. Peter Kemper estimated the overall cost of Sitescape Forum was around 600 per person per year. Much of this cost was human costs rather than IT costs with some 20 staff coordinating the sys- tem. He evaluated the cost savings from effective advice around the organisation at around 200m each year.4

On the question of Second Life, Peter Kemper was clear that a paradigm shift had occurred at Shell with a desire among users to explore new ways of creating and sharing knowledge. In terms of user needs, he continued:

'However, these needs are changing quickly. For example, these days, more and more users are familiar with Web 2.0 sites such as Facebook or SecondLife, and so we are also exploring the use of virtual worlds to add a new dimension to learn- ing at Shell - for example, with interactive 3D simulations and training modules. Shell Wiki has shown us that our users are ready to share their knowledge in new ways.'

Jeroen van der Veer's PA interrupted their discussions and informed them that the next meeting had been cancelled. Jeroen van der Veer suggested that they continue their discussions of virtual worlds and Second Life over lunch.

Questions

1 What benefits have wikis brought to Shell?

2 How would you promote greater user engagement with wikis in Shell?

3 What are the benefits of virtual worlds such as Second Life for Shell?