Journal of Knowledge Management Practice, May 2003

Knowledge Sharing And Organisational Learning:

The Impact Of Social Architecture At Ordnance Survey

Graeme Smith, Ordnance Survey, Dr. Deborah Blackman & Brian Good, University of Surrey


There are many debates at present as to how knowledge can be shared.  It is felt by some that there is an over-dependence upon IT based systems to share knowledge and that such systems do not reflect the complex nature of learning within social organisations.  The phrase 'social architecture' is being more commonly used as an element of the systems that affect and can support knowledge sharing.  This paper shows some of key dynamics that, despite a slowly growing awareness of their importance within academia, did not reach the discussion table of the project board implementing a knowledge management system. The paper indicates areas for future research and makes recommendations for future project management


Knowledge management is an increasingly important part of organisational strategy (Brint, 2001; Malhotra, 1998; Nasseri, 1996) and its potential benefits have been widely discussed (see for example Santosus and Surmacz, 2001; Prichard et al., 2000; von Krogh et al., 2000). This has led to senior managers in the UK public and private sectors striving to implement knowledge management strategies in their respective organisations.  Many of these initiatives result in failure, and very expensive ones at that. This paper will argue that this is because management has thought that the answer to the problems lay with the installation of an information technology solution. The implementation of a Customer Relationship Management (CRM) system as part of a knowledge management initiative is reviewed.  During the implementation process, it became apparent that the development was not going as well as expected.  Research determined that the organisational learning required to effect the behaviour change was not occurring, as a result of too little attention being given to existing social frameworks and architecture within the organization and too much to the technological changes required.

The overall aim of this paper is to discuss the role of social architecture in the successful sharing of knowledge within organisations. The impact of culture upon effective knowledge sharing is being increasingly accepted (Hall, 2001; Ill et al., 2000; Ryan, 1995) and the paper will explore further the role of assumptions, trust, reward systems, leadership, physical building and structural impacts and learning systems. The argument will be made that these latter elements of the knowledge sharing process have a far greater impact upon the potential to share knowledge successfully than any IT based knowledge management system currently has.


Ordnance Survey is Great Britain's National Mapping Agency.  It is widely regarded as the world leader in its field, which covers the production, maintenance and marketing of a wide range of maps, computer data and other geographical information for business, leisure, educational and administrative use.  Ordnance Survey has been a wholly civilian organisation since 1983; it is a free-standing UK Government Department and Executive Agency and, on 1 April 1999, it became a Trading Fund. This status allows it more commercial freedom than would otherwise be possible for a public-sector organisation.  Ordnance Survey’s core responsibility today is to survey maintain and keep up to date the national topographic database.

OS MasterMap is a massive database and online service derived from the national topographic database and lies at the heart of Ordnance Survey’s e-business strategy.  It includes seamless, polygonised data that is delivered online in themes such as roads, buildings and railways. Customers can also access change-only updates cutting down the amount of data they have to manage.

Unique 16-digit identifiers have been allocated to 416 million landscape features, including detail as fine as railway signal lights and free-standing letter boxes as well as every building and field.  These identifiers act as links between Ordnance Survey data and third party geo-demographic information, making it easy for users to compare, relate and analyse their information on a geographic basis.

It is updated on a daily basis by field staff using hand held computers to survey and digitise information.  Map data files of the survey area are called down from the central database on to the hand held computers and changes are made on the spot.  This revised data is sent back to the main Southampton database and processed overnight so the new information can be included in the database the day after surveying.

A huge range of products is created, using the information held in the database, and many services are provided.  For example, detailed information from the database can be accessed through channel intermediaries as hard copy printouts or as data on disc.  Direct supply remains the dominant channel to market for central and local government sectors and utilities.  Service level agreements (SLA’s) are seen as essential to the development of consistent, detailed product databases for these key sectors, as part of Ordnance Survey’s contribution to the present UK Government’s “joined up government” initiative, in essence the management and sharing of knowledge.

Why a Customer Relationship Management System?

The UK Government White Paper “Modernising Government”, sets out a vision for the future.  It is a blueprint for the biggest ever drive to revolutionise UK citizens’ dealings with UK Government.  As people have exercised choice and demanded higher-quality service in the private sector, service standards and service delivery have improved.  As a result of being able to shop from home, arrange mortgages without leaving the armchair and make holiday arrangements with one telephone call, increasingly the UK citizen wants the same standard from its public services and the White Paper sets out how this will happen.

Information technology is revolutionising business and the information age offers huge scope for organising government activities in new, innovative and improved ways of engaging the citizen/customer dialogue.  Steps are being taken to make more public services available 24 hours a day, 7 days a week.  By 2008, it will be as simple to do most of your dealings with UK Government as it is to make a phone call or switch between TV programmes.  In the future, the citizen will be able to authorise dealings with government by ‘signing‘ documents electronically using a smartcard.

The UK Government’s drive for greater synergy and efficiency across departments, evident in the “joined up Government” and “Modernising Government” initiatives, provides a major opportunity to demonstrate the true value of Ordnance Survey information as a common referencing tool for all areas of UK Government. E-Government initiatives impact on Ordnance Survey’s drive to satisfy a wide variety of needs through an increasingly sophisticated supply chain.  Not least of them is an increasingly discerning and demanding customer base, with increasingly sophisticated e-business needs.

Some key customers are looking for greater interaction in their customer/supplier relationship with Ordnance Survey.  They are no longer willing to trade with Ordnance Survey on a purely transactional basis.  As a direct result of a growing economy driven by e-business processes, customers are beginning to demand the same level of service that they receive from other suppliers, as well as reviewing that which they themselves provide.

Ordnance Survey recognises that if it is to deliver the vision of the White Paper it first needs to understand its own customer base.  It also needs to measure the real value of each customer, so that it can define levels of service appropriate to the needs of the customer and their value to Ordnance Survey.  Customer relationship management (CRM) was the means Ordnance Survey chose to identify, attract and retain customers, enabling an understanding of them and their interactions with Ordnance Survey and so Ordnance Survey invested 3.2m in CRM.  The initiative was designed to deliver a knowledge management system, which would ensure that customer knowledge is captured, co-ordinated and readily shared through the desktop. 

An analysis of business needs led Ordnance Survey to the belief that CRM practices would be highly relevant to its success for the following reasons:

        To achieve the core proposition of being the definitive supplier and co-ordinator of comprehensive mapping information for Great Britain for the benefit of all sectors of British life.  This requires differing levels of service and communication with a variety of suppliers, intermediaries, end users and those considered to have influence and this differentiation will need to be delivered consistently.

        Developing the geographic information market will require detailed customer information effectively stopping competitors causing market fragmentation.

        As costs of market supply rise with increased technology and channel complexity, so maintaining an optimum cost/income ratio is crucial.

        To support the Modernising Government aim of building consumer views on their needs into the planning process and improving the quality of service delivered as a result.

        Product innovation on its own is no longer sufficient to differentiate an organisation from it main competitors in an increasingly global economy.  To sustain long-term competitive advantage, organisations must seek to do more than simply bring an innovative product to market.  More sophisticated and discerning buyers seek added value from the relationship with the supplier, rather than simply trading on product and price.  Business customers are increasingly looking for suppliers who may help them to reduce costs through process alignment in their respective value chains.  This change predicates greater complexity in the relationship across the customer/supplier dyad.  CRM is seen by many as a means by which this relational complexity can be managed and, at the same time, maintain competitive advantage.

        To be successful a CRM system is dependent upon the access to information. Being a highly information driven organisation, such data was available both from within the organisation and from the customers.

Thus a CRM strategy and implementation programme was put together to build up Ordnance Survey’s customer management capabilities, absorbing sales, service and marketing processes and developing them within an overall customer management framework. The CRM system was seen as the conduit for knowledge management within Ordnance Survey, as the only way that CRM could be effective was if there was the application of the information held and generated by the new software by ‘knowledge workers’ (Starbuck, 1992; Andrews and Delahaye, 1999).

What was the problem?

The choice of a CRM system for Ordnance Survey appeared to be eminently appropriate. At all levels within the organisation the plan was proposed, discussed and accepted.  However, as the project developed it became apparent that there were stumbling blocks with the implementation and the expected changes in outputs were not emerging.  The problems included: a lower level of staff awareness than had been anticipated, not enough of the people who should have been involved in the desired outputs were, the organisational culture was still resistant to the changes and many people were ignoring new systems they considered as irrelevant to them.  The result was that the desired knowledge management system was not being developed.

Ordnance Survey was interested to know why this should be and so research was undertaken looking at what had been expected to occur by developing and installing the new system, and then comparing this with what was actually emerging.  Interviews were undertaken with key personnel, meeting minutes were examined for the decision processes and the implementation program, training and communications were considered. All of this data was analysed for themes and patterns. It emerged that there were three key assumptions being made within the organisation which underpinned the way the project was conceived, planned, communicated and implemented. 


Assumption 1: System implementation changes processes and behaviour

As was outlined earlier, the reason for implementing the new CRM system was to provide a basis for knowledge that would enable individuals within the organisation to behave in a different way so that their reactions to customers and situations would alter.  This desire for different behaviours is key to the current situation because implicit within this is a need for learning to take place (Levine, 2001).  A commonly cited definition of learning is a ‘relatively permanent change in behaviour’ (see for example Robbins, 1989 p.62), indicating that before there can be a change in behaviour there must be learning taking place.  Therefore, if the new systems and processes are not developing new behaviour, then the inference is that the new knowledge management systems are not enabling the learning needed in order to trigger the desired new behaviours.

From discussions within Ordnance Survey it became apparent that it was thought that, by changing the work processes within the organisation in order to facilitate a sharing of knowledge, then new behaviours would emerge as a direct result.  Thus the new knowledge that was accessible to all would encourage learning and new behaviour.  That this was a reasonable belief can be seen when considering this statement from Argyris and Schon:

“Generically an organisation may be said to learn when it acquires information (knowledge, understanding, know-how, techniques or practices) of any kind by whatever means.  In this overarching sense, all organisations learn, for good or ill, whenever they add to their store of information, and there is no stricture on how the addition may occur” (1996, p.3).

There is a suggestion here that merely enabling the provision of more information is in itself learning.  Several writers have indicated that there is confusion between information and knowledge in general terms within organisations (Blackman, 2001), and this too may be leading to a greater focus on the provision of new information, rather than upon the processes that develop new knowledge and behaviours.

However, learning, be it organisational or individual learning, is not dependent solely upon wider access to new knowledge.  Argyris and Schon (1996) are very clear that organisational learning will result in something that is different – having more information alone is not learning, nor will it engender learning. Merely having a knowledge management system, and providing access to it, will not produce changes of behaviour or lead to greater understanding. Effective knowledge management only occurs if there is a knowledge sharing culture which promotes learning.

In order to learn, there needs to be a motivation to learn (see for example Gagne and Driscoll, 1988) and this was where a part of the problem was occurring.  The individuals within the system had access to the new information but, in order for them to use it and to develop new ideas, they needed to see why or how it would be of use to them.  This awareness should lead to a predisposition to both learn and develop its use, but there was no such awareness within Ordnance Survey. Despite the training and communications used within the company to implement the project, the focus had been upon what the system did and how to use it; there was no picture of why it would advantage individuals. 

There is another important issue here, the current behaviour of employees should act as an indication of the level of learning. In Ordnance Survey, the view taken was that as there were no overt problems then things must be going well. However, it seems that the view should have been that if the behaviour does not change then the learning is static also.  This was a very clear level of feedback that needed to be considered more carefully – especially where no change was occurring.

Assumption 2: Systems are mechanistic in nature

Morgan (1986) and Daskalaki (2000) both indicate that there are many ways of looking at organisational realities.  One way is the utilisation of metaphors.  In this case, in discussions with employees (especially managers) within Ordnance Survey, there was a mechanistic view of the organisation as a machine, with the employees as cogs working within it.  The focus within the company upon business process re-engineering and the way it is being described supports this view of the organisation as a machine that can be rebuilt, fixed and have new parts fitted.  Such a metaphor meant that it was assumed that changing the machine mechanics (i.e. organisational systems and processes) would naturally change the outputs in some way.

Assumption 3: hat the organisational systems will be rational

There was an assumption of logical cause and effect relationships between the inputs and the outputs of the systems and processes.  However, for this to occur there needs to be a rational approach to knowledge generation; this would mean that there is a set of definite truths which can be accessed by all (Russell, 1961). Rationalists state that a core set of knowledge can be deducted by reasoning in a logical way.  Within knowledge management development, this would indicate a set of systems and processes that could be determined to be the right way to manage the organisation in order to gain the desired outputs; the search for the appropriate input will lead to the right output.

However, there is an increasingly accepted view that much knowledge within organisations is constructed by the individuals working within it (Easterby-Smith 1997; Easterby-Smith et al., 1999; Evans and Easterby-Smith, 2000; Blackman, 2001).  It is logical, therefore, that if there is to be the requisite organisational learning, required to develop new behaviours, there must be a recognition of the constructed nature of knowledge within the new system development.  This is currently being reflected within knowledge management literature.  McAdam and McCreedy (1999) indicate that many knowledge management models reflect the old managerial paradigm (as outlined by Clegg et al., 1996), which does not recognise the importance of the individual within the system (see figure 1, black arrows are expected behaviour, and white arrows represent actual behaviour). 



Figure 1: Demerest's Knowledge Management Model

McAdam and McCreedy (1999) go on to show that to ignore the constructed and social nature of knowledge development is to oversimplify the behaviour modelling and, therefore, to underestimate what needs to be done to develop knowledge management systems.  They develop a new model that reflects this greater level of complexity (see figure 2).

Kakihara and Sorenson agree with this picture of complexity in the development of knowledge: “knowledge, is by nature emergent in terms of its interpretative, process-oriented and relational properties.  And knowledge in organisational contexts is generated through complex, dynamic interactions between actors, organisations and social environments” (Kakihara and Sorenson, 2001 p.16)

This dynamic property of knowledge development is becoming increasingly accepted (Carter et al., 2001).  Evans and Easterby-Smith (2000) state that “organizational knowledge creation is less a transformation or amplification of individual and group knowledge and more a result of productive inquiry consistent with the act of ‘knowing’”. Whilst Cook and Brown (1999) talk of a ‘generative dance’ where new ideas are interpreted via several input sources within the organisational context bringing forth new meaning and new knowledge.  Both papers demonstrate an iterative process where new knowledge develops from reactions with current knowledge and new ideas.  They also reflect a common view that within the system is tacit knowledge held within individuals.

These elements to knowledge creation infer that there is a current level of knowledge within an individual that can be developed and interacted with, in order to develop something new again.  Unless the individual engages with the process, what they already know cannot be either ascertained or accessed, thus it will be outside the system.  It was apparent that the level of current understanding within Ordnance Survey was acting as the basis for the future, because no learning had taken place and so the workers were not knowledgeable. As the employees could see no reason why they should change their ways of working they were not engaging with the implementation process, thus the possibilities of the CRM system were not being achieved.

A Communications Problem

Three assumptions have been outlined above and together they may begin to explain why Ordnance Survey was having problems implementing CRM.  The researchers chose to consider Knowledge Management at Ordnance Survey as if it was a communication system, in order to discover whether using this metaphor explained why learning was not occurring as expected.  The choice of a communication metaphor was made due to the increasing focus upon the need for incremental understanding within the system.  This is a key element of any communications model (see Robbins, 1989 for example) and it was hoped that this differing perspective would help diagnose the problem more clearly.

From the first assumption, it was clear that many messages were being sent but that not enough consideration was being given to the feedback within the system.  Because the system was seen to be the driver, the output was being assumed and not enough thought had gone into ensuring whether inter-mediation had occurred.  Many new processes had been developed in order to implement the new system, but the reasons for the new behaviour were not clear to those undertaking them and so no changes were actually occurring. Moreover, the assumption was made that undertaking the process was enough for there to be change. However, because to be a knowledge management system there needs to be knowledgeable workers, it proved possible to apply the processes with no useful changes in customer relationships.

The second assumption again led to the ‘buy-in’ of the individuals within the system not being considered as vital to the process.  If one were communicating at this point, how the message would reach the receiver would have to be considered.  This debate was not occurring as far as knowledge management was concerned.  Because the people were seen as being merely elements of the system, the messages were not being received or understood. The last assumption meant that messages were almost certainly being encoded wrongly, as the constructed nature of the knowledge being developed by individuals was being ignored.  What was apparently happening can be seen in figure3:

Within Ordnance Survey there was a great deal of input as regards systems and new ideas but not enough ownership and interest from those it should have been affecting.  Simon (1991) argues that unless the bounded rationality of both individuals and organisations is developed in order for the new knowledge to be recognised and accepted, then the new knowledge may be present within the system as a whole but it will fail to make an impact.  At the time of the project implementation at Ordnance Survey several factors were preventing the development of the new ideas: the culture was not encouraging knowledge sharing, the training had not communicated a clear need to individuals and there was no change in the reward and performance management that led individuals to believe that the new system actually mattered to them. These factors led to individuals filtering out the new information as unimportant.

The argument was made that in this case too much expected learning was being based on cognitive learning theory, where an abstract explanation (as in sending round a memo or setting up a procedure manual) is used.  Individuals are expected to find out for themselves but, if they have no interest or motivation, this is unlikely to work and no learning will occur.  Elkjaer (2000) advocates a change in focus on organisational learning to concentrate more upon active learning techniques.  Such an approach is being encouraged within the education system, as it encourages engagement in the process (Zemke and Zemke, 1995), and it seems logical that similar ideas can be utilised to encourage organisational learning.  Vital for the success of such an approach is a clear understanding of the learners and this fits with the model outlined above.

Figure 4 shows a different approach to communication that identifies it as learning and shows where the behaviour changes will occur.  What can be seen is that the vital element for the success of the model is to ensure that the encoding of the message is done in such a way as to promote and encourage understanding.  It is the contention of this paper that, for this to be done effectively, the social architecture of the host organisation will need to be evaluated and understood.

Was the problem due to ignoring Social Architecture?

Originally the term ‘social architecture’ was coined as a form of architecture intended for use by the mass of people as social beings and was a reaction against architecture concerned with form and style supposedly for the dominant members of society (English Heritage, 2001).  Increasingly, it is now being seen as very important when discussing the design and relationships within organisations (see for example General Electric, 2001), when stressing the role of community in architecture (Batteau, 1996) and when designing new organisational structures (Jacobs, 2001).

Bennis (1997) considers social architecture when discussing the requirements for effective self-managing teams, implying the relationships within the organisations will potentially support or destroy the teams.  Kelley (1987) employs the term when discussing the issues of connectivity within organisations.  He outlines social architecture as the social system which is one of the five key areas of connectivity that can be managed, in order to improve the IT effectiveness within the organisation.  Morden (1997) uses the term when determining the elements of visionary leadership.  He outlines Kay’s (1993) definition of architecture as being the network of relational contacts within and around the enterprise, since these relationships will affect both trust and communication.

It was the idea of social architecture affecting interpersonal relationships, structures, leadership, communication and, therefore, the successful passing on (or not) of messages that was of interest to the researchers in Ordnance Survey.  Knowledge Management is fundamentally about transferring information in such a way that the user can both understand and use the meaning of the transfer in some way (Elkjaer, 2000; Klein and Myers, 1999).  It has been established that the real Knowledge Management problem being faced by Ordnance Survey is a communications one.  If knowledge management is essentially a communication system, then logically it must be affected by the social architecture of the organisation.

What is important is that an awareness of social architecture may permit an organisation to attain a level of self-analysis that is not being achieved at present.  The real advantage appears to be recognition of the issues that are affecting the outputs of the system.  An analysis of the current situation led to a model of issues currently affecting the implementation of knowledge management within Ordnance Survey.  Some might argue that this is an analysis of the ‘soft’ systems within the organisation (Checkland, 1999; Mabey et al., 1998) and to a certain extent this is true.  There is a network of relationships which are affected by filters such as feelings and individual interest at all times (see figure 5).

All communications models show filtering and ‘background noise’ as key elements in the success (or not) of message and understanding transfer.  By rethinking learning as being framed by the social architecture, it can be seen that it will be acting as a filtering system.  For example, emotions at work have always been acknowledged to have an impact but, increasingly, it is seen that they will seriously affect the effectiveness of the outputs of any system (Clutterbuck and Megginson, 1999; Weisinger, 2000).  They will always affect the potential receiver and will change over time.  All of the elements of figure 5 could be seen within Ordnance Survey but, as has already been indicated in this paper, several could be seen to have had substantial impacts upon the implementation process and its lack of success.


There were two areas which, had an analysis of Ordnance Survey been undertaken, would have demonstrated reasons for a lack of trust that was fundamentally undermining the project. Firstly, after 11 months of planning the implementation of one CRM software solution, there had been a decision taken by the board to change to an alternative provider. This provider was already being used within the business for back office solutions and so it seemed logical to senior management to use some of the front office functionality. Two questions were raised by this decision: why the original choice and why change now? The first choice had been chosen by the project board as it was considered to be “best of breed” and the most suitable solution. The decision to change this was part of a perceived need to implement an enterprise wide software solution for data delivery, ordering, stock control, general ledger and financial account management as well as CRM, in an attempt to integrate the front and the back office. It is clear that the decision was driven by the technology not necessarily the outcomes desired by the project. To change after such a time clearly undermined the respect for, and trust in, the project team and challenged the importance of CRM as anything other than a set of software processes, especially as experience showed that the front office functionality of the new system did not enable coherent delivery of CRM processes. Employees observed processes being “shoe-horned” into the software capability and questioned the validity of the implementation process and the expected outputs, which again challenged the project team’s apparent ability to deliver.

Secondly, at the time of the project, there had been a re-organisation and a new management regime in place for six months.  There was a perception that no obvious benefits had been delivered and employees were beginning to lack confidence in the effectiveness of the new management.  Within the e-business strategy there were many other initiatives, and some viewed CRM as one initiative too far.  No re-evaluation of the CRM implementation strategy, in line with the restructuring of the organisation and its new business strategy, took place until there were obvious problems. It was apparent that many people did not see why they should change.  The concerns felt by the employees led to a level of mistrust that would act as a filter against the new ideas, as they could not perceive the advantages CRM would bring.


The culture at Ordnance Survey was variously described as traditional, old-fashioned and obedient. This led to the employees adopting new processes in a passive, apparently accepting way. However, as outlined above, this did not mean that they accepted the ideas in such a way as to lead to learning and change. CRM had been developed in the private sector and was widely used and understood in this context. Such acceptance was not prevalent within the public sector. In fact, CRM was being viewed by the management as a way to become more like the private sector and change the culture of Ordnance Survey.

This lack of real understanding meant that the current organisational culture viewed the implementation with suspicion. The obedient nature of the organisation meant that some instructions were followed, but the knowledge needed for CRM to work and develop new outputs did not emerge as there was no acceptance of the private sector ethos. In other, less consenting, cultures the lack of real acceptance might have become apparent earlier. The type of culture also led (as indicated earlier) to management assuming that their choice of communication systems would work as no feedback was equated to positive acceptance.

Reward Systems

If there is to be behaviour change there needs to be a motivation for this and the change must be rewarded in some way. People will generally act in a way that they perceive as being rewarded – this is not merely pay but outcomes that will make an individual feel that they are achieving some of their intrinsic or extrinsic needs (Palardy, 1994; Mullins, 2002). In this case, no efforts were made to reward the actual outputs anticipated as being directly attributable to the CRM system and so only the observable processes were adopted. This can be seen to be because of the assumption of an organisation as a mechanistic process with the inputs being the key focus; it was assumed that if the employees undertook the processes the desired outputs would automatically emerge.

Currently, the organisation is implementing a Performance Related Pay system which will be related to Return on Investment (ROI), which is hoped to act as a major incentive within the company. However, unless the employees see CRM as directly increasing ROI  and, therefore, change their behaviours in order to make it work, such a reward scheme will fail to support the successful implementation of CRM. At present there is no evidence that the new reward scheme will encourage the implementation of CRM. In fact, should the bonus be received without CRM working as the project team had hoped, this will reconfirm old behaviours and make successful implementation even less likely.


Owing to its origins as a mapping factory, the leadership of Ordnance Survey was historically hierarchical, centralised and transactional. Although in recent years there have been efforts to change this, the implementation methods were, for the most part, reflections of the original leadership style, with the decision to implement CRM being taken by the directors upon the submission of a business case, itself developed by a project team looking at how production could be improved through more sales and better efficiencies. The case was based upon benchmarked information gathered from the private sector by the team. This looked good and was accepted by the organisational management team but, in hindsight, it is clear that there was top down acceptance but not top down commitment and understanding: as one director put it “we heard what you had to say and we listened, but we couldn’t feel, touch or smell it”. This led to a lack of obvious leadership for the project which led, in turn, to employees adapting to orders, but not necessarily actually understanding the CRM system and not perceiving the project as being important enough to actively learn about it. This created an apparent compliance but led to the failure of successful implementation.

Learning Systems

The historical learning systems at Ordnance Survey focused upon formal inputs and upon clear instructions. This equates to the acquisition of information rather than the development and use of knowledge. Where such processes are prevalent, a lack of understanding often develops and original processes may re-emerge over a period of time unless there is enough re-iteration and support to ground the changes (Senior, 1997; Hatch, 1997). This is what has occurred within Ordnance Survey as, even after the system has been in use for nearly a year, most of the sales staff do not use the full account management functionality. They know what to do but they consider themselves to be too busy to use it – it is perceived as slow and cumbersome to navigate and as not providing them with tangible returns. Repeating training will not solve these problems, only if sales employees can see how using the system will improve their target achievement will they change. No learning system was in place to deliver the ‘why’s of the system, only ever the ‘how’s.


A further difficulty was the way that business cases were used for making all major decisions. Effective learning is partly dependent upon the quality of the information that is available. It is widely acknowledged that the form of the decision making processes will affect the output of such decisions (Brunsson, 1989; Dutton and Ashford, 1993, Evans, 2000). Furthermore, the characteristics of the decision makers will affect the types and results of the decision taken. 

The Ordnance Survey quality system documentation gives a detailed representation of the formal expectations of the decision making process within the organisation, including the formation of a business case. The preparation and content of the business case demonstrates the collection of evidence, the formulation of options and the means of identifying preferred options by (financial) cost/benefit analysis.  The process describes a linear progression through business case approval to production and launch.  Within the assumptions noted in the business case, the presumption is that Ordnance Survey can expect to achieve the predicted revenues and benefits outlined in the approved business case. However, there are two problems here. The first, outlined above, is that information is likely to be limited in its accuracy. The second is that the characteristics of the individuals involved in the case will decide how the case is presented and the likely outcome.

The original business case was sold on the back of a great deal of hype about CRM originating from the industry and financial institutions. As a result, despite the business case being extremely weak, it was seen by senior management as a ‘must have’ and was therefore accepted. Further weight was given to the case as the proposer was an individual who had implemented a similar system in an organisation Ordnance Survey benchmarked against. This led to implementation from a basis of weakness (a poor case) and the people who should be involved were largely forgotten in the push to achieve. The organisational and IT inputs were the focus, and not the potential gains.

At this stage Ordnance Survey was pushing very hard to become a successful e-business. Despite its potential as an implementing role for all the e-business developments, the CRM implementation was separated from other e-business strategies because of the management structure implemented. In order to give each project the power base that it needed to be implemented quickly within Ordnance Survey, each project (of which CRM was one) had its own separate board which managed its implementation and reported to an overall e-business strategy board which in turn reported to the main board. This led to an unexpected reversion to hierarchical structures and powers which led to the isolation of CRM from the rest of the business. As a result there was little learning across the projects and the knowledge needed to successfully develop the CRM system was not being developed as fast as expected, particularly as the project proposer was not appointed as the project manager. The focus upon CRM as a people-centred solution was lost and, partly because the director responsible was an IT consultant brought in specially, the hardware solutions became the primary locus of power.

Building Design

Ordnance Survey is located in a large 1960’s style building which was primarily designed for production. The building is shaped as an E and so the open plan offices are very long and thin. To visit others in the building is a long walk and so most communication is done via e-mail and telephone. Work teams are usually co-located and when the CRM team was formed it was put together. However, the location of the team within the building was a reflection of the separation of the structure and the power streams. Upon its formation, the CRM implementation team was not co-located with the other e–business teams or customer services, which was devolved as an ‘add on’ to other functions, thus no informal relationships could be built. This lay out enabled learning and knowledge within the team, but prevented an open dialogue and opportunities for learning across the projects and for developing the knowledge required to implement the project into the areas where it was designed for use.

From analysis of the elements of the Social Architecture, as applied to the implementation of CRM in Ordnance Survey, it is clear that there were many filters to learning – people were concerned about their futures, felt a lack of trust and were demotivated. As a result, employees could not see the relevance and, therefore, lost confidence in the initiative. Where decisions were taken, they can be seen to have been based upon (a) the assumptions in place outlined earlier and (b) the software and technological solutions. It seems likely that, if a broader analysis had occurred, some decisions would have been different.

Because the company had not explored or researched what new encoding was needed to enable knowledgeable workers, no potential understanding was developed and no learning occurred.  Figure 6 shows a potential model for setting up new learning reflecting the importance of social architecture, which might have prevented the implementation problems at Ordnance Survey.

A new metaphor should be used so that managers no longer think of the organisation as a machine.  Marshak (1993) indicates that changing the root metaphor used for organisational analysis enables a broader understanding of the social systems within the organisation, thereby promoting change.  By using a changed metaphor of learning as communication, and by thinking of the organisation as a network of “social architecture” the managers at Ordnance Survey may rethink the way they set up implementation systems.  They will, hopefully, be more output driven and be more aware that unless the employees are enabled to be knowledgeable workers no changes can occur.


This paper has shown that many knowledge management implementation problems are about the way the processes and systems for knowledge creation are being developed in isolation from the social systems of the organisation.  Rather than being a process problem, poor knowledge emergence from a new system is more likely to be a communication and learning problem where there is a failure to engage with the individuals who are within the system.

However, at present many of these issues do not seem to be taken as seriously as they need to be within the organisations themselves.  By framing a model of social architecture for a specific organisation, it should be possible to work out the best way to engage the individuals within the system, in order to begin to communicate in a way that will trigger learning and, therefore, behaviour change.  By spending more time on organisational diagnosis, benefits will accrue sooner, money will be saved and the pain outlined in this case study avoided. If such an analysis had occurred prior to, or even during, the implementation of CRM into Ordnance Survey, the knowledge outputs and organisational effectiveness changes desired would have occurred considerably earlier.


Andrews, K.M. and Delahaye, B., 1999, 'Determinants of Knowledge-Sharing in Organizational Learning: the psychosocial filter', 3rd International Conference of Organizational Learning, June, Lancaster University, <<http//>>.

Argyris, C. and Schon, D.A., 1996, Organizational Learning II: Theory, Method and Practice, Massachusetts, USA: Addison-Wesley.

Batteau, A.W. 1996, ‘The Social Architecture of Community Computing’,

Bennis, W., 1997, ‘An interview with Warren Bennis’, Training, August, 14. 8; 33-37.

Blackman, D.A., 2001, ‘Is Knowledge Acquisition and Transfer Possible?’, Electronic Journal of Radical Organizational Theory, 7. 1; Vol7_1articles/blackman.asp.

Brint, 2001, ‘Knowledge Management and Organizational Learning’, BRINT Web Site,

Brunsson, N., 1989, The Organization of Hypocrisy, London: Wiley.

Carter, C., Mueller, F. and Swan, J., 2001, 'Social Capital and the role of HR practices in the management of knowledge’, Conversations and Critiques, Managing Knowledge Conference, April, University of Leicester.

Checkland, P., 1999, Soft systems methodology in action, Chichester : Wiley.

Clegg, S., Barrett, M., Clark, T., Dwyer, L., Gray, J., Kemp, S. and Marceau, J., 1996, ‘Management knowledge for the future: innovation, embryos and new paradigms’. In Clegg, S. and Palmer, G., (Eds.), 1999,  The Politics of Management Knowledge, London: Sage.

Clutterbuck, D. and Megginson, D., 1999, ‘Emotional Intelligence in Executive Mentoring’, Plenary Paper at 6th European Mentoring Conference, November 25th-26th, Cambridge University.

Cook, S.D.N. and Brown, J.S., 1999, 'Bridging Epistemologies: The Generative Dance Between Organizational Knowledge and Organizational Knowing', Organization Science, 10.4; 381-400.

Daskalaki, M., 2000, 'Re-assessing the use of metaphors in Organisational Settings: the boundaries of Induction Experience', BAM 2000, September, University of Edinburgh.

Dutton, J. and Ashford, S., 1993, 'Selling issues to top management', Academy of Management Review, 18, 3, 397-439.

Easterby-Smith, M., 1997, 'Disciplines of Organizational Learning: Contributions and Critiques', Human Relations, 50.9;1085-1113.

Easterby-Smith, M., Burgoyne, J. and Araujo, L., (Eds.), 1999, Organizational Learning and the Learning Organization, London: Sage.

Elkjaer, B., 2000, 'Learning and getting to know: the case of knowledge workers', Human Resource Development International, 3.3;343-360.

English Heritage, 2001, ‘Social Architecture, Thesaurus, styles/c178164.htm

Evans. I., 2000. ‘The role of business cases in decision making process of Ordnance Survey’. MA Marketing Masters Thesis, Southampton Business School.

Evans, N. and Easterby-Smith, M., 2000, 'The Nature of Organizational Knowledge', The British Academy of Management Conference, University of Edinburgh, September 13th-15th.

Gagne, R.M. and Driscoll M.P., 1988, Essentials of Learning for Instruction, (2nd Ed.), Englewood Cliffs, NJ: Prentice-Hall.

General Electric, 2001, ‘Social Architecture’, architecture.htm

Hall, B.P., 2001, 'Values development and learning organizations', Journal of Knowledge Management, 5.1;19-32.

Hatch, M.J., 1997, Organization Theory, Oxford: Oxford University Press.

Ill, P.E., Kessler, E.H. and Christenses, E.W., 2000, 'Organizational Learning, Knowledge and Wisdom', Journal of Organizational Change Management, 13.6;595-618.

Jacobs, R.M., 2001, ‘Management and leadership as creative “social architecture”’, The Structure and Theory of Organization, /Powerpoint/f-structural/Index.htm.

Kakihara, M. and Sorenson, C., 2001, ‘Exploring Knowledge Emergence’, Conference on Managing Knowledge: Conversations and Critiques, 10th April, University of Leicester.

Kay, J., 1993, Foundations of Corporate Success, Oxford: Oxford University Press.

Kelley, R.E., 1987, ‘Raising Corporate America’s Connectivity Consciousness’, PC Week, September 29, 4.39;C24.

Klein, H.K. and Myers, M.D., 1999, 'A set of principles for conducting and evaluating interpretive field studies in information systems', MIS Quarterly, 23.1;67-93.

Levine, L., 2001, ‘Integrating Knowledge and Processes in a Learning Organization’, Information Systems Management, Winter, 18.1;21-33.

Mabey, C., Salamann, G. and Storey, J., 1998, Strategic human resource management: a reader, London : Sage.

Malhotra, Y., 1998, ‘Knowledge Management for the New World of Business’, Asian Strategy Leadership Institute Review, 6.Special Issue on Knowledge Management.

Marshak, R.J., 1993, 'Managing the Metaphors of Change', Organizational Dynamics, 22;44-56.

McAdam, R. and McCreedy, S., 1999, ‘A Critical Review of Knowledge Management Methods’, The Learning Organization, 6.3;91-100.

Morden, T., 1997, ‘Leadership as Vision’, Management Decision, 35.9-10;664-673.

Morgan, G., 1986, Images of Organization, Sage.

Mullins, L., 2002, Management and Organisational Behaviour, 6th Edition, FT Prentice Hall: London.

Nasseri, T., 1996, Knowledge Leverage: The Ultimate Advantage, BRINT, papers/submit/nasseri.htm.

Palardy, M.J., 1994, ‘Principles of Learning: a review’, Journal of Instructional Psychology, 21.4;308-12.

Prichard, C., Hull, R., Chumer, M. and Willmott, H., (Eds), 2000, Managing knowledge: critical investigations of work and learning , Basingstoke : Palgrave.

Robbins, S.P., 1989, Organizational Behaviour: concepts, controversies and applications, 4th edition, Englewood Cliffs, NJ: Prentice-Hall.

Russell, B., 1961, History of Western Philosophy, London: Routledge.

Ryan, M., 1995, 'Human resource management and the politics of knowledge: linking the essential knowledge base of the organization to strategic decision making', Leadership and Organization Development Journal,  16.5;3-10.

Santosus, M. and Surmacz, J., 2001, ‘The ABC’s of Knowledge Management’, CIO Magazine, May 23, html# benefits.

Senior, B., 1997, Organisational Change, Pitman Publishing.

Simon, H.A., 1991, 'Bounded Rationality and Organizational Learning', Organization Science, 2.1;125-134.

Starbuck, W., 1992, 'Learning by Knowledge-Intensive Firms', Journal of Management Studies, 26.6;713-740

von Krogh, G., Ichijo, K. and Nonaka, I., 2000, Enabling knowledge creation: how to unlock the mystery of tacit knowledge and release the power of innovation, Oxford : OUP.

Weisinger, H., 2000, Emotional Intelligence at Work, San Francisco, Ca: Jossey Bass.

Zemke, R. and Zemke, S., 1995, 'Adult Learning: what do we know for sure?' Training, 32.6;31-37.

Meet the authors:

Graeme C. M. Smith, Business Improvement Manager, Ordnance Survey; Graeme has worked for Ordnance Survey for 29 years, starting in surveying and cartography and then moving into Marketing. He was the Manager of the CRM project and is now managing Customer Services.

Dr. Deborah A. Blackman, Lecturer in Human Resource Management, University of Surrey; After working in the hospitality industry Deborah moved into academia. She teaches, researches and consults in the areas of human resources management and change management, with a particular interest in the areas of organisational learning and the impacts of mental models.

Brian Good, Management Tutor, University of Surrey; Initially Brian followed a career in the Health Service. He then moved into teaching management, specialising in human resources. He now runs the Virtual MBA at the University of Surrey.