Feature

posted 1 Jan 2001 in Volume 4 Issue 4

Just-in-time knowledge

In a relatively short space of time, and with limited resources, Smith & Nephew has succeeded in creating a knowledge-centric intranet for international marketing and sales staff within the company. Simon Hudson explains how the tool has developed, and offers his advise on what traps to avoid, and which elements of the system have worked well for the pharmaceutical giant.

It seems that, in the information age, we are faced with something of a paradox. On one hand, as managers and as companies we are expected to be more effective, produce better results and use no more resources than before. On the other hand, the amount we are required to ‘know’ to achieve this exceeds practical learning limits. Time spent learning could be spent doing.

If only we could access the information we need, when we need it, we could improve our efficiency and effectiveness and so deliver the continuing growth that is expected.

We are entering a phase of ‘just in time knowledge’. Training is no longer a sustainable instrument to keep up with developments.

Fortunately the problem comes with the germ of a solution, in the form of knowledge management as a discipline and its embodiment in various personal processes, many of which may be technology-enabled.

Drivers and issues

In our particular case, as a large, many tiered global company, we faced our own ‘unique’ set of problems (unique only in that they are set against the background of the particular structure and culture of our organisation):

Large amounts of data and information exist within the organisation

Much of the desired information is isolated

Country offices do not have access to full information to support them

Speed of access to information is low

Much effort is spent recreating or re-searching for information

Key information sources are people, many of whom are highly mobile personnel, often out of office for weeks at a time.

A survey within the international marketing department in early 1999 revealed that information requests from our country markets took up to 10 days to complete; 20 per cent of queries also went unanswered even though the required knowledge was known to exist in the organisation.

In addition, we had limited intelligence gathering, analysis and dissemination, with no defined process and limited reporting structures. As a consequence, management often did not have easy access to the information if needed to support good decision-making.

It was these issues that we hoped to address through the introduction of knowledge management approaches, the most visible being the creation of a global intranet.

Objective

We started the process by defining what we hoped to achieve. The intent was to:

Share information and market intelligence

Make information available when and where needed

Simplify and accelerate access to information

Improve the mechanism for associating items of information to allow patterns to emerge and hence new inferences to be drawn

Improve robustness of the company’s intellectual capital (so that it could be retained irrespective of people leaving or travelling).

This objective was set against a common set of constraints and hindrances:

Limited resources

Low level of executive buy-in/interest, largely due to the breadth of other important projects concurrent with this one

Limited technical infrastructure (two prototype intranets existed, one was limited to a single geographical site in content and reach and had minimal uptake; the other had some good content, but was pitched at a very general level and was not being driven throughout the organisation).

Concerns and opportunities

There were several issues we had to face in establishing the knowledge management approach.

Firstly, we had to rethink the usual approach to managing information as a consequence of the nature of information we hoped to store. We had attempted, in the past, to store such information in the classic relational database. This assumes that the information to be stored and the relationships between the types of information can be predefined. This traditional IT approach had failed, as the types of information we wished to store had elaborate and chaotic relationships with other items of information - it defied pre-emptive classification. In this respect I tend to think of this as being knowledge, rather than simply information. This issue was summarised in a colleague’s observation that "IT is a large part of the problem, but only a small part of the solution".

Another issue was trying to find a balance between the three key aspects of knowledge management, especially where intranets are concerned:

It isn’t just about technology (though it is very seductive to many of the people involved)

It isn’t just about the information itself (though some people make a living selling it)

It isn’t simply about what people know (knowledge that is not applied and built on is of minimal value unless you are entering a quiz).

It is about using technology to allow people to apply knowledge to meet business needs.

We had concerns about ongoing knowledge management and especially document management. How could we ensure that the desired information was being captured, evaluated and eliminated once its usefulness had expired? How could we create global access and use in a fragmented infrastructure and organisation?

What we did

Initially we undertook various forms of research. Most significantly we attended a knowledge management conference, organised by Frost & Sullivan. This proved to be of immense value, exposing many questions and issues and providing an opportunity to create the mind space that we used to build our approach. A number of key areas of learning stood out from the conference:

There were many more questions than answers - even among those companies with successful intranets.

The approach to content validation on intranets was divided into two camps: Those that had an ‘open door’ policy - i.e. little or no content validation; and those who used a system of ‘gatekeepers’ - i.e. all content passed through the hands of one of more individuals for approval, before being published on an intranet. The latter approach produced slow content approval and high stress levels among the gate-keeping individuals.

The system for document management on intranets also fell into two camps: Those who applied an arbitrary and draconian document management policy (such as deleting all documents over two years old); and those who had no policy at all. The latter case usually turned out to be a delayed form of the first option, as an arbitrary mechanism for controlling the number of documents tended to be applied once the intranet became too large.

I will address our solution to the second and third points later in this article.

We also spent time discussing what else had been attempted within our organisation, particularly within our research company and with the people who had been involved in the existing intranets.

Finally, we assessed what type of knowledge and information we held within our part (International Sales and Marketing) of the organisation, which elements would lend themselves to storage on an intranet, and which areas to prioritise based on a combination of value to the business and ease of creation. This gave us a valuable working list, which provided some quick wins and some high value wins. It also allowed detailed discussion with our IT department to establish a potential technology platform for meeting these needs and identifying a capital cost and project plan.

This learning was set out in a white paper, which was circulated for comment and widely presented to the department and other potential stakeholders. Within this we identified a number of important aspects of our approach:

We intended to shift the communication paradigm from push to pull (although we have now moderated this to provide a balance between the two). This can be considered to be changing the culture from ‘know it all’ to ‘know it when...’ and would reduce ‘death by email’. (I was receiving upwards of 70 emails per day at this point, so was somewhat motivated by this issue.)

We intended to create an intranet as a central business communication tool, with the IT systems that support it being treated as core business systems.

Everyone is personally responsible for their own data, with no central gatekeepers.

By default, all published information would be available to everyone in the company (though restricted access to documents would be supported when necessary).

We were lucky enough to have a professional Lotus Notes developer on contract. He was central to our progress, both by developing the applications themselves and also by helping to define the art of the possible (and continues to be a driving force for the evolution of our intranet strategy). On the basis of this and many other factors we selected Lotus Notes and its Domino web server application as our technology platform. In particular, the free-form nature of Notes lent itself to storing the rich, document-centric information our research had suggested was needed.

I attended an intensive one-on-one Notes training course. This allowed me to understand the abilities and constraints of the application, improving the ability for our developer and I to work together. Ultimately, I was able to do most of the user interface design and maintenance while our developer created the core capabilities of the intranet applications.

Significantly, we did not follow traditional IT project planning, implementation and documentation processes. As what we were trying to achieve was rather different to any project completed by IT, it was agreed that a more freewheeling approach was valid. Consequently our development cycles were very short, typically of the order of 2-4 weeks for first release, with upgrades occurring on the fly. This is very much the paradigm we see today in Internet developments, enabled by the server-thin client model (in that only the application on the server needs updating for everyone to have access to the new functionality). Now that the intranet is established we are introducing more rigorous controls, as users have come to depend on a stable, recognisable intranet.

Our intranet is built on a number of discrete document databases, including the following components:

Knowledge-base - This was conceived as a general ‘document’ store, holding news, marketing information, competitor information, general product information, multimedia, presentations and virtually any type of content not stored in a dedicated database. The content of this database is provided by everyone/anyone in the organisation. Any document is given one or more types of categorisation by the author when it is entered. Consequently, various documents are able to be associated with each other in different views. Thus a document about a particular new product launched by a competitor to tackle a specific clinical problem could appear in a section to do with that competitor, in a section on new products or in a section about that clinical problem, even though the document was only entered once. The assumption is that different users have different needs and approaches to browsing for information. The different views or sections were designed to cater to these needs, allowing a richer means of creating associations or identifying patterns between otherwise isolated pieces of information. It is also important to note that we set the knowledge-base up so that only the author and anyone he or she nominates is able to change their document. However, anyone is able to create a response document to an existing document, expanding it, critiquing it or connecting it to other documents.

Yellow Pages - This is our people database, with a photograph, location and responsibility information, areas of knowledge and skill, contact information, plus a brief biography. It is really useful for new starters who need to put names to faces, and also is a primary means of finding telephone numbers and other contact information across our extensive international organisation. Individuals maintain their own entries, though we created most entries for them using information gathered at various marketing meetings and via email.

Dictionary/thesaurus - This started life as a simple acronyms database, yet it has evolved into an extensive dictionary of business and clinical jargon, complete with images and links to related terms.

Market research library and clinical papers library - These two components hold title, author, publisher and other reference information, plus abstracts and/or tables of contents for all the physical reports and papers we hold. Increasingly, we purchase reports electronically and therefore publish the whole document online. User permissions are used to manage any copyright restrictions. We also purchase a number of external reference databases of market information from Espicom Ltd, which are replicated regularly from their servers to ours.

Marketing materials - We maintain an extensive range of marketing give-aways, products and related literature, plus product and clinical photography. All these are now available to browse and order online. The image library is especially useful as it allows PowerPoint size images to be downloaded across the intranet. We have disallowed transfer of the print ready versions of the images, due to their very large size (20Mb+), but these can be requested on CD ROM, while the smaller downloadable version is adequate for print agency mock ups.

What works well

A few thoughts on what has worked well for us:

We addressed the issue of gatekeepers by eliminating them. We have placed the responsibility for the content on the author/publisher of the information. This has been maintained by a combination of training, visibility and peer review (the author of the document is visible on every document) plus the introduction of a reliability rating, assigned by the author, which indicates how robust the information is. Although this approach has met with some resistance (we are a highly regulated industry, so there is an expectation that we have independent quality controls on most things), we have experienced very few incidents where poor information has been posted. Where this has occurred the ability to immediately update the information (without recalling hundreds of emails or paper copies) has proved effective. Nevertheless, we are introducing digital signatures and may introduce a structured gatekeeping concept in future, (although the gatekeeper will be the author in most cases!).

The issue of document management has been tackled by the introduction of a document ageing method. This requires that the author assign an ageing rate value to each document (between 0.25 and 60 months), which works in conjunction with the reliability value to give the document an individual document lifetime. When a document expires it can be deleted, archived or referred to someone for review. The default is to delete the document. Typically, news items (for instance) have a reliability of 3 and an ageing rate of 2 months, and are set to delete on expiration - these documents are automatically removed from the database after six months.

We have found it best to give users what they need rather than what they want. Don’t spend too much time deciding on the intranet functionality by committee. Our experience is that there was no correlation between the things users said they wanted and their desire to use them if available. Instead, we used our knowledge of our co-workers daily needs as a guide to prioritising content. We then promote and monitor usage.

We carefully monitored the balance between push and pull. Initially we tried to eliminate most of the push behaviours, but we discovered that users liked receiving an email to prompt them to do something or alert them to the existence of information. Our compromise has been to email links to the information they may want. This reduces the amount of data travelling over the email network and encourages users to look at other related documents while using the intranet to retrieve their linked document.

From the outset it was intended that there must be multiple ways to find/access any piece of information. Initially it was expected that the various views that Lotus Notes allows would offer sufficient ease of access, but we found that, beyond about a thousand documents, this method began to break down. To counter this we have added a top-level search engine, which is able to search all (or a user selection) of the databases that drive the site, plus an automatically generated document reference number (it is now possible to simply advise someone to look for document KBS1946, for instance). We also introduced the concept of the SIP, a Special Interest Portal, which is one or more pages built by a user, hosting links to documents, databases or views relevant to their area of responsibility; a product manager, for instance, can create a mini-site for their product. The main site portal has, to date, been revised four times to accommodate the increasing complexity of the content.

Pitfalls and failures

There have been a number of things that have not worked well for us:

Our use of Notes plus browser is not ideal, as users have to use two interfaces; a browser to retrieve information and Notes to enter it. Ideally, all interactions would be through the browser, although it has not proved possible to enter rich (formatted) text, multimedia, graphics and multiple attachments through the browser. This will be addressed with future technologies.

Users are confused about having to log into the browser, after they have already logged on to the network.

Discussion groups have simply not been used by our group, despite a good understanding of their potential value. Email remains the preferred tool for non-real-time communication.

Branding was an issue for us; we did not establish a strong intranet name/brand, with the result that users are confused between the portal, the intranet, Snip (our high level corporate intranet) and the knowledge-base.

Traps we have managed to avoid are:

Poor ease of use. We have been consistent in the look and feel of all the areas of the intranet, but the complexity of the information we are storing continues to challenge our ability to design good navigation methods.

Planning for perfection. We didn’t expect a perfect system. Instead we have allowed room for evolution; things which work well initially will degrade as the knowledge system expands, while other approaches that are unnecessary or unworkable to begin with become viable later. Don’t expect every facility you provide to be used by the users, even when they have requested that feature. Also don’t expect everyone to use it - there will be an adoption curve (classic Gaussian distribution in our case).

Deferring investment until business or IT conditions are ‘more favourable’. It is so hard to provide a good P&L argument for managing these intellectual assets. Our initial intranet development was inexpensive enough that we have been able to justify it on instinct, build it, learn from it, use it and can now re-engineer it based on that learning.

It is extremely easy to publish copyright protected materials to an intranet, especially by copying items from the Internet. We have adopted a practice of using a short abstract and a link to the source document on the Internet.

Beware the 5 Syndromes:

‘Not invented here’ syndrome

‘It’s your database’ syndrome

‘It was good enough for my father and my father’s father’ syndrome

‘I haven’t got time’ syndrome

‘It’s not my job’ syndrome.

Conclusion

We have, in a short space of time and with limited resources, been able to build a sophisticated intranet tool capable of storing many types of corporate knowledge. This tool supports many of the activities undertaken by the marketing teams around the globe, allowing our teams to spend more time acting and less time learning or requesting information. In the process we have learned much about the complexities of the knowledge we deal with daily, exposing potential new processes for improving our effectiveness and increasing the demand for similar (and better) solutions elsewhere in the business. Where previously intranets were below the radar screens of IT and most of the executives, they are now being pursued as important business tools that enable the front end of the business; this attitude change is a reflection of our corporate shift towards a knowledge managing culture.

Finally, don’t underestimate the effort involved in changing the culture of your organisation. We expect it will take two years before the intranet and knowledge management are unconsciously integrated into daily practice, but the signs are excellent that we will achieve this.

Simon Hudson is e-marketing and strategic intelligence manager at Smith & Nephew. He can be contacted at: simon.hudson@smith-nephew.com