If only we knew what we know

Henkel Group's association with knowledge management was triggered by a desire to tap into the know-how residing in the minds of its 60 000 employees. Myles Marcovitch describes Henkel's experiences and outlines the organisation's global approach to KM.

The situation

"If we only knew what we know..." Is this the rallying cry of the latest management fad knowledge management? Coming on strong at the end of the 20th century KM brings together the notion that competitive advantage is getting harder to come by through technical innovation alone and that information technology is at the fingertips of almost every employee in a modern firm. Started in the big multi-national management consulting firms and a few farsighted enterprises KM has taken on the life of a full-fledged movement within global businesses of which Henkel is a prime example. But KM has its difficulties. For one thing there are many definitions of knowledge management and like other major performance movements preceding it such as Total Quality to grasp the concept is easy to put it into sustainable action is not. Nevertheless it also holds great promise.

Henkel being a large international corporation with 60 000 employees worldwide realised that it too would be much stronger if the know-how residing in those 60 000 minds could be shared more effectively. Henkel is successful by any measure: In its fifth generation of family ownership with a turnover of €12 billion and in strong market positions in many regions around the world the firm has always been profitable but things can always be better. Market innovations although giving some competitive advantage are more and more transient as competitors react ever more quickly. However organisational innovations that energise people are much harder to copy and can provide long-lived market strength. And KM can help do that.

To effectively implement a KM process at Henkel we required a little outside help. The company retained the services of Arthur D. Little (ADL) as the prime contractor for this project. In autumn 1998 ADL presented their version of KM at Henkel's senior management summit and the decision was made to proceed with a limited foray into the world of knowledge management. ADL had been using this process for some time and had operational experience with it as have other international consulting firms. Since knowledge is their primary product it makes sense that formalised KM processes have evolved there first.

KM is taking two main forms in organisations. The first uses database techniques to capture and store information in the form of knowledge pieces that can then be retrieved by users who wish to learn from the knowledge contained within. The other seeks to create organisational structures and interactions that promote people-to-people sharing of know-how. Henkel has both of these processes at work today but the latter is a topic for another article. This discussion focuses primarily on the database form of KM.

Simply speaking the function of a KM process is to decode tacit knowledge (that which resides in our heads) in order to make it explicit knowledge (that which can be put on paper and shared with others) and then re-encode this explicit knowledge to allow others to use it for their own success. Throughout our formal education we were indoctrinated that copying is a form of cheating; as such we have to overcome many established behaviour patterns.

ADL's process centres on a series of structured interviews called debriefings. It is through this process that know-how residing in the minds of individuals is decoded and put in a form that is storable and retrievable by others. This is not a simple process for debriefings are not just a means of capturing history and stories. They must capture lessons learned from successful and unsuccessful events and do so in a form that is both complete and interesting. We also recognise that it is impossible to capture all know-how through written means. I can read all of the books I want about Tiger Woods but to play like him...

You need full-time resources to properly control this process and ensure its quality. In the ADL world debriefings are conducted at the end of each consulting project or milestone. Use of the system is included in their consultant's performance management system as well as part of the culture of performance. People build career strength through the numbers of debriefing abstracts on the system that bear their name. Here at Henkel the use of the system is much less organised. But as it grows we will be expecting greater application in people's daily work.

Getting into action

Our KM experience began by selecting five pilot projects: Two consumer two industrial and one support services. Each project was chosen based on the leverage it could exert on the company's success if know-how were shared at a much higher rate than previously practised. Each project had a team leader appointed by a senior-level steering committee that oversees implementation strategy. The project leaders make up a core team that meets regularly to make tactical decisions that ensure successful project completion.

Pilot projects moved quickly with phase one ending in June 1999. Completion of the pilot phase was determined by a set number of debrief abstracts being put in to the system and the organisational structure set in place to support the KM process in the future. As a result of the early success these five pilots have been extended to encompass broader scope and depth and four more projects have been added.

Features

Our KM model consists of five parts; these are interdependent and each must be robust and implemented effectively (see figure 1).

Figure 1 - Henkel's knowledge management model

Content is the know-how/knowledge pieces brought to the system by subject matter experts (sometimes called SMEs) - the givers.
Process is the means to decode tacit knowledge to make it explicit knowledge - the debriefing sessions - and to disseminate it throughout Henkel.
Context is the framework through which knowledge pieces can be located and retrieved easily - using a keyword thesaurus structure.
Culture is how we all behave in Henkel which either supports or resists the sharing of know-how across our many boundaries. Culture is influenced by what is rewarded or punished in an organisation. Are people encouraged to share what they know or are they asked to develop ideas alone?
And lastly infrastructure is the K-base system that stores and enables us to retrieve the knowledge pieces easily.

Implementation steps

Based on key business processes the first activity in implementation (after forming the teams of course) is to define the key business process steps and sub-steps that comprise the primary search fields in the database. All debriefs are tied to these business process steps. It is also important for the teams to prioritise the process steps. There may not be enough time - or even the inclination - to document every step in a process. You only need to debrief on the process steps that make a difference in the success or failure of the outcome of the business process.

As an example the HR recruiting process looks as it appears in figure 2. Steps one to four were considered critical and are therefore primary debriefing targets. The remaining steps will be debriefed if time permits. Step seven is after the new hire begins work and is probably the beginning of the next HR process either 'retention' or 'employee development' and will be covered at a later stage in the KM rollout.

Figure 2 - the HR recruiting high-level process flow

Once the main process steps are defined key sub-steps are identified and prioritised. These categories and sub-categories are directly installed into the database as fixed themes and are used as keyword search fields. Within each sub-category the team identifies subject matter experts (SMEs) who have relevant experience to share. 'Debriefers' are assigned from the team and a schedule put together to conduct the debriefing sessions.

It is not enough to interview people and get the story; the debriefer is also responsible to ensure that the abstract he or she has created answers the important 'so what?' question. Debrief abstracts when read by another employee (possibly a novice) must convey enough know-how to allow the reader to act on the information. This is not easy to do and often requires the debriefer to go back to the SME for a second review to ensure that the nuances are captured. For this reason we do not recommend that people insert their own abstracts into the system. Critically reviewing your own abstract to ensure all of the details are present is extremely difficult. It is much better to have an objective third party conduct the interview and ask the 'naïve' questions of the uninitiated. Only then can the underlying know-how be deduced. This is also a learning experience for the SME. You would be surprised how often improvements in the business process can be designed just by asking basic questions of 'why' and 'how'.

Recently I was privileged to conduct a debriefing of Ulrich Lehner Henkel's new CEO. The knowledge managers realised that we didn't just want casual support from top management; we needed their full commitment. What better way to gain their attention then to have them directly experience the KM process through a debriefing session? This debriefing session is being inserted into all of the KM databases for everyone to see.

Putting the information into the Lotus Notes database is not difficult particularly if the abstract is created in an MS Word form specifically designed for this purpose. It is merely a cut and paste task once the abstract is edited. Keywords are associated with the abstract when loading it into the system; these are accessed by pull-down menus for fixed keywords or are added by the debriefer for modifiable keywords.

One of the KM process' most challenging aspects is keeping the system alive and up-to-date. It is the job of the knowledge manager to set up periodic reviews of all abstracts in the system and cull those that are no longer valuable. He or she must also ensure that document links are still viable and point to the correct location.

Each person who contributes information to the system is forever linked to that information and is asked to ensure that when they move locations and jobs they notify the system of the change in details. Otherwise in a very short time links to documents and SMEs are lost. There is a button on each knowledge piece that enables the reader to instantly email the contributor for additional information on the topic. There is also a commitment by contributors that they will be available for additional questions on their abstract.

User access is a vital part of the entire KM process. Without users taking know-how out of the system applying it to their own work and realising faster success KM is an exercise in futility. Therefore we take user training very seriously and have online and paper versions of the KM user training programme and KM debriefer training.

Roles

Different roles have been mentioned in this paper; perhaps it is best to define them here.

Senior supporter. Provides direction removes roadblocks frees up resources and facilitates the implementation of KM.
Knowledge manager. Full-time resource at the business unit level who manages the process on a daily basis ensures the quality of the abstracts in the system coordinates system maintenance selects trains and coaches debriefers and measures KM performance.
Debriefer. Full or part-time person who conducts debriefing sessions edits debrief abstracts inputs them into the k-base and trains and coaches users.
Users. Everyone who seeks to gain know-how more quickly by using the KM process.
IT support. Information services people who maintain the technical integrity of the KM system.

Lessons learned

Our experience helps identify a number of factors that can ensure a more successful implementation and operation of a KM system. These include:

It is vital to ensure the quality not just the quantity of debriefs in the system. There is great temptation to just 'get the system loaded'. Avoid this temptation! The first time a knowledge user enters the system and reads one weak abstract after another will be the last time they use the system.
Focus efforts only on business processes perceived to have high leverage on business performance. There will never be enough time or energy to load all of the know-how in a company as large as Henkel into a KM system nor would anybody ever want to. Make sure you prioritise the business processes to be explored and the sub-steps within them.
Measure key success factors in business performance terms. If this process is driven by the businesses it has a much greater chance for long-term success than if it were driven by staff groups like HR or information services. Therefore all measures must be made in business terms based on business results. Of course you will want to measure the number of 'hits' on the system as an activity gauge but what we are really interested in are the results of applying this new-found know-how to generate more volume market share and profit.
Frame benefits in terms of personal benefits to those who use the KM system. Users require some additional motivation to use and update the KM system that transcends key business results. The system must also create personal success by being able to do things faster more successfully with lower stress than before.
Ensure full-time participation of key resources especially in the pilot phase. Another temptation to avoid is wedging the KM creation and management process into already busy jobs. Much of the work involved with bringing KM into an organisation is labour intensive. It is a significant change effort requiring many meetings presentations and interactions. KM also requires the strict attention to detail associated with computer systems. Being a knowledge manager in a business unit can be a full-time job!
Use effective change management techniques. Communicate often and garner high-level support that backs its words with deeds and resources. Installing KM is no different than any other organisational change task. There are soft and hard issues that must be addressed through communications interactions and coaching. All roles in the KM process have a part to play in making this change happen. Getting sufficient resources is one sure way to tell that senior level support is real. HR plays an important role here.
Train everyone who interacts with the system. Don't expect that users will intuitively know how to frame search questions when seeking know-how from the system. We are relying on the debriefers as the primary coach to help users at their worksites. Debriefers also receive training custom-designed by HR due to the unique nature of the debriefing interview. We will be using an online package to teach the basics of knowledge management and thereby reduce in-class time requirements.
Ensure participation represents the broadest constituency and is multi-national whenever possible. Henkel's KM implementation is worldwide; therefore the teams are multi-national in makeup. For example the HR team consists of three Germans three Americans two Spaniards one Belgian one Frenchman and one Chinese. Debriefing sessions are being held in all of these locations. The language in the system is English but depending on the subject matter expert the interview could take place in the native language and then be translated. Of course this complicates the process and can lead to possible misinterpretation of specific details.
Stay with an infrastructure that you already know. We decided to stay with our familiar Lotus Notes application. This means that people all over the Henkel world familiar with the Lotus Notes interface will not need additional training on Lotus Notes only KM's special features.
Resist the temptation to this make this an information technology 'gee whiz' project. Many knowledge management initiatives originate in the IT group. After all the data warehouse concept seems to be what KM is all about but not in our interpretation. Because of the desire to capture know-how and lessons learned and not just data KM must be led by people throughout the organisation. There is certainly an IT component as we are storing the abstracts in an information database but this is not the focus of the system. IT plays an important support role but it is not the driver of the project. We have been using our own lessons learned and simplifying the system to continue to improve its user-friendliness.
Recognise that this is an on-going process without apparent end. It must be maintained to keep the knowledge fresh and relevant. Know-how is volatile. It ages sometimes like a fine wine other times like raw meat. The goal of our KM process is not simply to load the system in one big bang and then declare success and move on to another business initiative. It is rather to create a sustainable new way of sharing know-how into the future.

Overcoming resistance to implementing KM

Our experience indicates that getting people to contribute has not actually been that difficult. Since we are still in the early phases of implementation and have not asked people to contribute time and time again there is still some novelty in contributing what you know and helping others. Each of the businesses has specific KM goals concerning the minimum number of debriefs that should be completed by project and business teams. According to ADL the culture is evolving to the point that for employees to have their name on a number of abstracts is considered 'good' while not being seen in this context is 'bad'. It is our belief though yet untested that as knowledge users extract valuable know-how out of the system they will also be positively inclined to want to put their know-how into the system. It will not be seen as good behaviour to only be a 'taker' and not a 'giver'.

One of the most positive aspects of the Henkel/ADL model is having the expert's name associated with the knowledge piece. This implies a certain 'immortality' where people are long-remembered after they are no longer with the particular job they were doing when they were debriefed. People respond positively to this concept. It is very important to emphasise that having your name on an abstract in the KM system obligates the giver to be a coach and advisor to others who seek more understanding than what is written on the page. It is a physical impossibility to capture all knowledge but with the experts identified there is a second chance to obtain a deeper understanding. Since our approach is a global process these experts could come from anywhere in the world and under normal circumstances would never have the opportunity to share what they know.

Since all knowledge pieces are tied to keywords and to the experts that gave them it is a very simple matter to generate an expert 'yellow pages'. One of the search options does just that. An expert yellow pages is one of the most basic forms of KM. In our system it comes along with the basic installation.

The most important means for encouraging people to use the KM process is to have an organised plan to conduct debriefs at critical project milestones. The implementation team originally sets these policies. After gaining experience the knowledge manager fine-tunes the process to set a workable and pragmatic debriefing schedule. The last thing you want is to force people to get debriefed at the end of every activity. Quality goes down and people quickly tire of the paper chase. It is better to have open conversations about what constitutes a 'debriefable' activity based on sensible criteria that pay off in high quality know-how in the system. It is important for the knowledge manager and the debriefers to know which projects are on-going within their areas of influence and which are nearing a milestone and as such a debriefing point. It is also important that they know who is being challenged and would benefit from knowledge within the system. As the process matures people will come forward and volunteer to undergo a debriefing when something of interest happens. Managers need to point people in the right direction and encourage them to access the KM system at the beginning of projects so that they conserve the time spent planning.

We do not on the other hand have a great deal of information on resistance to using the KM system. We anticipate that people will find great value in it and therefore will spread the benefits through word of mouth. We are actively developing internal marketing tools to encourage use through knowledge fairs articles adding knowledge sharing to our list of management core competencies and awards programmes. This awards programme - presently in the design phase - features awards for users based on the success gained from application of know-how from the KM system. Additional awards of the same value will also be given to the knowledge giver and debriefer of the same knowledge piece.

The future

We are still in a growth process even though we are coming up on two years. Many of the issues surrounding KM implementation revolve around culture change and this can easily take three to five years to realise. Also the debriefing process is labour intensive and takes a considerable amount of time. We have good data about people putting knowledge into the system. What remains to be seen is the level of activity in taking knowledge out. We are very serious about our commitment to this process and expect that it will return value greatly in excess of the investment. Early results show people reacting favourably to what they are finding in the system and we are getting comments like 'If I only had this when I started project X two years ago'. If these feelings persist knowledge management will be a part of Henkel for a long time to come.

Myles J. Marcovitch is chief learning officer of the Henkel Group. He can be contacted at: myles.marcovitch@denotes.henkel.de