The collaborative advantage

Managing content to enable knowledge sharing

One of the ways to ensure that a content management solution supports and enables knowledge sharing is to have as much content as possible voluntarily home grown by the community being served by the solution. Dan Holtshouse shares his experiences at Xerox, where he has helped to pioneer and develop this concept.

Retention of organisational knowledge and expertise is a continuous problem in both good and bad times. That’s why many companies have knowledge sharing and content management as top business priorities in their firms and why they are using knowledge sharing initiatives involving communities of practice to build and retain expertise.

However, there is a human resistance to sharing in most business environments. When people see their knowledge as a source of job security or power they are often reluctant to share it, especially in a business climate of hard economic times. Consequently, companies need to create a work environment with a culture and incentives that are conducive to sharing, and then support that environment with improved work processes and strong technology.

It’s incredibly important in a knowledge-driven environment that people are encouraged to use other people’s ideas. The hard reality is that the world we are living in is moving at too fast a pace to allow repeated re-invention. Individuals have value in a knowledge economy if they know something and are willing to share it, or if they can find and use something that someone else knows. Regardless of tenure or experience, people who are contributors will have worth in the knowledge-driven company; those who are seen as hoarders will be left out.

Key to systematic capture and retention of personal and community knowledge is creating a dynamic flow among the community members and the content repositories. Based on our experience with several large knowledge sharing initiatives, this is best realised by building a knowledge environment where the community itself creates a significant amount of the content. To avoid the problems of content not being used, we have consciously chosen to build each content management system in and around the community it serves. This requires a socio/techno balance to the design of the knowledge sharing environment and the content management system that supports it. It requires assessment of the community needs for content and an access plan for the content that is integral with community work process. In our experience, one of the keys for successful content management is to have as much of the content as possible voluntarily home grown by the community itself.

When the community helps build content, many problems are solved. The content is more likely to be relevant because it comes from activity generated within the user community itself. Comprehension problems are reduced because community members converse in an agreed to set of terms, definitions, customs, symbols, etc. Trust is enhanced because authoring accountability and peer recognition imply certain standards and quality levels for the contributions for continual good standing in the community. So, if the content is created by the community for the community, and is readily accessible within the community, we believe knowledge sharing will happen naturally.

What follows are four case studies illustrating this process of balancing the socio/technical to achieve dynamic and active co-building of the content base in order to ensure maximum knowledge sharing. These case studies cover several large communities with differing needs, including customer service, sales and marketing, software engineering, and call centre operations. But the processes of achieving active co-building of the content base is similar across these otherwise different communities.

Knowledge sharing in customer service (Eureka)

The first project is a global knowledge sharing system in the customer service organisation of Xerox aimed at improving customer satisfaction, reducing service costs and building systematic capture, retention and re-use of new knowledge. Known as the Eureka System for the capture and sharing of solution tips for newly experienced product problems, the system grows community knowledge by encouraging customer service engineers (CSEs) to voluntarily submit new solutions to problems not seen before (so that fellow CSEs will not have to re-invent the same solution somewhere else across the globe). Each tip is authored by name and validated by an experienced colleague to ensure quality entries that will be acted upon by other CSEs looking for a solution.

The knowledge base containing these valuable tips for new problems is maintained on a central server and is downloaded to the CSEs laptop regularly so it is available for use at a customer site visit along with electronic documentation, training materials and on board diagnostic tools.

Everything a CSE needs is on the laptop hard drive because access to a customer line is often problematical while at a customer site. Downloads and updates are usually made a few times a week when the CSE is back at the Xerox office.

When doing a search of the Eureka knowledge base, the CSE utilises a completely tailored workspace that reflects the way a CSE searches for solutions to problems. The user interface (UI) is:

A browser metaphor design, usable whether the CSE is online or not. The browser was chosen because we learnt that CSEs used the internet on off-hours so UI training would be minimal;
Customisable by information category, ie, by software, hardware, service, validator, etc;
Customisable by personal workflow preference, ie, by product, problem or symptom, etc;

Once a problem solution is selected, a wealth of information pops up giving community-based insight on the author, validator and how it’s been rated by others with a thumbs up/down score. Since all tips are validated by a process known to the CSEs, they know they can trust the knowledge-base content and can act on it to solve a customer’s problem while on site.

The embedded work symbols of the CSE community in the Eureka system represent no mean accomplishment. To design the ideal system, Palo Alto Research Center (PARC) computer scientists, behaviour scientists and anthropologists travelled around with the CSEs in France for many months in the mid-1990s to discover how they worked, what motivated them and how they shared war stories within the workgroup.

The flexible design of the Eureka system, the fast search engines, the browser-like UI, the novel motivation ideas around solution authorship and the validation process were the result of many months of trial and error by the PARC scientists. Eureka is, in many ways, the result of an artful integration of people, process and technology.

From the early experimentation in France in the mid-90s, Eureka has been scaled up into a global knowledge sharing system supporting over 25,000 CSEs worldwide in seven languages. There are over 50,000 solutions in the knowledge base – all submitted by the CSEs on a voluntary basis. No extra time was given for solution entry into the system.

Eureka is credited with solving over 350,000 problems in 2001 that otherwise would have been addressed by other CSEs, wasting both parts and labour as they try to recreate a solution. Parts and labour savings are in excess of $15m annually, in addition to increased customer satisfaction and faster learning cycles for the CSEs. Eureka, we believe, is a good example of dynamic content created by the community itself, resulting in a high value knowledge sharing environment.

Knowledge sharing solutions for sales and marketing

Aspects of the Eureka learning experience, system design methodology, software tools and community activation techniques are being transplanted to other Xerox communities such as sales (GrapeVine) and marketing (Focus 500). These solutions are designed to manage customer relationships and accelerate sales cycles – one by tracking, reporting and sharing knowledge about our largest accounts, the other by helping Xerox people reproduce successful customer solutions in other locations around the world.

Major account management (Focus 500)

Focus 500 can be thought of as an inexpensive customer relationship management (CRM) system that enables Focus executives, national account managers and global account managers to track, report and share knowledge about calls made by Xerox executives on the company’s top 500 accounts. Focus 500 is a client-server application that works on the Xerox intranet. The information in the database is downloaded to the user’s hard drive every time they sign on, ensuring that users have the most up-to-date information and that the software can respond quickly to search inquiries.

The repository houses a directory of all executives participating in the Focus executive programme; an account-information repository replete with customer profiles and summaries, and links to other information provided by account managers; and a ‘who do you know?’ section listing the names of Xerox executives who may have door-opening contacts at specific customers.

Every time a Focus executive makes a customer call, they enter information into the system about the call: who made it, who was spoken to or visited, when the call was made, the type of call, a summary of what was discussed and a description of any new business opportunities that were identified. A series of information templates are used, designed on the basis of a series of interviews with users, to enable rapid capture of input.

Both local profiles and global profiles of customers are kept in the system by regional offices so that executives are well prepared for meetings with customers. In just a few months, the system has captured several hundred customer visits across the globe, all voluntarily logged into the system. Senior executives, account managers, product support specialists, etc, are all co-building the content and the knowledge base together.

Customer application solutions (GrapeVine)

GrapeVine shares many similarities with Focus 500; its technology is identical, although its application and target users differ. Whereas Focus 500 supports high-level executives, GrapeVine was designed to support primarily those who sell and develop Xerox solutions, and the people who manage them.

Partly because of the mobility of its users, GrapeVine is delivered over the internet. Users can connect via the Xerox network or by dial-up connection from anywhere in the world; data is encrypted to provide security.

The intent behind GrapeVine is to promote the re-purposing of solutions, thereby increasing revenue, reducing the length of the sales cycle and lowering costs by improving sales productivity. The heart of GrapeVine is a growing collection of solution successes that number in the hundreds, each of which was voluntarily contributed by a sales and marketing specialist with a rich description of a successful solution and/or a competitive win.

The solutions contain a wealth of information about each success. This includes the customer’s name, industry sector and market segment, the type of solution, the decision maker’s title, the business problem addressed, a description of the solution and its benefits, the products and services used, the revenue generated, the competitive face-off and whether the win represented new business. The customer solution content is being voluntarily built by sales and marketing because they will have access to information and tools to better and more quickly target customers and propose proven solutions. Field and headquarters managers will also save time in managing the sales process, and solution developers will gain insights into which solutions are selling; into where, how and to whom they’re being sold; and into new solution requirements. The knowledge base, although still less than a year old, already contains several hundred solutions, many of which were newly developed for the first time at customer sites.

Software code sharing (CodeX)

Software code production is a growing component of engineering development in the Xerox line of products. To become more effective in code creation and more efficient in code re-use, a code sharing project called CodeX was created less than a year ago. To get started, a grassroots champion for the project modified some public-domain code and built a web-based service to help engineers swap software. The Code Exchange initiative (CodeX) has grown to include over 900 registered users and is home to more than 100 projects, including some of the company’s largest cross-organisational software efforts.

The research scientist from the Xerox Grenoble Lab in France who co-founded the project went around the company talking about the benefits of software sharing and found that there were few practices of software re-use, let alone software sharing. Even within workgroups, people weren’t using code they’d written the year before.

It was soon discovered that a long-standing barrier to code sharing has been a lack of means. There was no way by which software developers could track projects outside their domains or search for existing code whose re-use might save them work. So the project team set out to create one, taking advantage of development techniques from the worldwide open source software movement, which features free code exchange, peer contribution recognition and early exposure to end-users.

The CodeX team had its initial website running by January 2001 after 90 days of development. Several iterations later, it has become a destination where developers can browse through more than 100 software projects categorised by type of software, development status, intended audience, operating system or programming language. Since December, the most popular categories were general software, printing and internet/intranet connectivity, scientific engineering, information technology, document management and image management.

In addition to posting and searching for code, CodeX users can hold topic-related discussions, exchange messages, pose questions and offer solutions to the Xerox-wide software development community, plus follow links to related Xerox and external open source projects. They can also find a toolkit with which they can host and manage entire development projects on CodeX. The kit includes high quality public domain (and therefore free) version control and bug tracking software. After moving its software development work to CodeX, the 90-person Xerox Architecture Center found that it could save half a million dollars in software license fees.

Within a year of the website opening, content contributed by the software community has built up from zero to a complete array of re-usable tools including version control, file release, web server hosting, task manager, bug tracker, document manager, web forum, patch manager, peripheral driver and a range of applications software for printing and internet connectivity.

CodeX still has a way to go before it tips the scale toward institutionalisation. With 25 to 35 per cent of the software community active co-builders, and growing, the software content base is now approaching ten million lines of code. Yet it has become a respected company-wide content site for re-usable software and a marketplace for the exchange of ideas among developers. It’s also becoming a foundation for efforts to instil a new knowledge sharing culture among software developers, to improve time-to-market and product inter-changeability, and to lower the cost of developing Xerox products.

Knowledge sharing in telephone call centres (Angelo)

Customer call centres are notoriously busy, noisy and stressful workplaces. Employee burnout and turnover are high as operators cope with large call volumes, demanding customers and the need to scour many sources for constantly changing data on complex new products, services and technologies.

Recently Xerox has worked on a project with five European partners to ease working conditions while improving customer service at Telecom Italia (TI). Like many of its counterparts, TI faces aggressive new competitors in a post-monopoly environment, while coping with the demands of digital-age services and technologies. As a result, it’s trying to retain customers and earn their loyalty by providing high quality service. Alongside handling typical questions about accounts, pricing, coverage, contracts and billing, customer service reps in TI’s call centres now field inquiries about how to configure and troubleshoot ISDN lines and ADSL modems, mobile phones and internet services. Information about these digital services is also constantly changing.

Before implementing the project, called Angelo, operators used at least three different databases and many different steps to find the electronic information needed to answer customers’ questions. In many cases they had to wait as long as 90 seconds for an answer to a query. As a result, many operators kept paper cheat sheets and folders at their desks to keep track of shortcuts to the electronic information and paper-based product and service updates.

Through the Angelo project, disparate information sources along with a new dynamic knowledge base are now integrated into a single interface. It has received rave reviews from its users in pilot tests and is based on the core methodologies, experiences and tools (LinkLite) of Xerox’s Eureka – a peer-to-peer knowledge sharing solution.

The system offers its users a single interface on their PCs where they can find consistent and up-to-date information in a new knowledge repository of problem-solving tips as well as in various existing intranet databases and legacy mainframe systems. The system also lets operators initiate telephone calls, manage and search e-mail, hold instant chat sessions with supervisors, share applications with colleagues and contribute their own tips.

Data about customers and their accounts, available phone numbers and other administrative information is stored in a mainframe-based legacy system. Information about company products, services, calling rates, etc, is stored on the intranet. The new knowledge repository contains troubleshooting tips as well as shortcuts to content in the other databases. Memos and product, service and administrative updates, normally distributed on paper, are captured in Angelo and made available through LinkLite. It’s a combination of process and technology to capture and share community knowledge as well as structured information.

To design the system a research team from PARC, including ergonomists, ethnographers and cognitive psychologists, interviewed call centre operators and observed their work practices. They learnt where operators looked for answers to particular questions, how long it took them to find an answer, when they used electronic or paper sources, and when they called in a supervisor. The work practice observations and interviews served as a basis for the development of new work processes, technologies and incentives for their use. To contribute a new tip, they need only click the button labelled ‘proposing a new tip’. This is sent to a supervisor for validation and is added to the LinkLite database. During the project pilot phase operators contributed over 250 tips alone.

As we have found, the Angelo project discovered that there were two primary incentives for operators’ use of and contribution to the project: the peer recognition that came with attachment of contributors’ names to their validated tips; and the improved problem-solving ability that came with faster access to higher quality, up-to-date information.

Operators like the flexible, single-point access to multiple information sources, the quality and timeliness of the knowledge in LinkLite, the near immediate response to their inquiries (it was more than three times faster than TI’s intranet and legacy systems), and the ease with which they could co-build the tip knowledge base.

In a nutshell

In these four case studies, I have attempted to make the argument that successful content management is as much about getting valuable content in the content base as it is about building the content infrastructure. It is about balancing the social/technical dimension to ensure tight work process integration with the community it serves. It is also about the dynamic formation of community action so that the community members themselves take on an active role in voluntarily building the content.

I believe that by co-building the content base, familiar community language is ensured, relevant content is provided, and high quality knowledge is created through validation procedures and authoring processes that can be trusted. Relevant, timely, trusted content enables successful knowledge sharing. By building content management systems from the community out, with the active participation of the community we have the best chance of getting people to bring the best of themselves to their jobs and to willingly share their most valuable asset – their knowledge.

Dan Holtshouse is director, corporate business strategy at Xerox. He can be contacted at dan.holthouse@usa.xerox.com.