back Stories Overview

User Experience at SAP



Conference Reports

Future Scope


Knowledge and Innovation in Administrative Areas – Part I

By Dr. Carlos Bezos Daleske, responsible for Corporate Development at DaimlerChrysler España Holding – September 14, 2004

The following article describes a PAR (participatory action research) experience in a business environment (DaimlerChrysler, Spain), where the research method has been strongly ethnographic. It was divided in two parts for easier reading. Part I provides an introduction to the field, presents the methodology used, and reports a case study.



Traditional administrative areas have undergone some dramatic changes since the first wave of SAP rollouts in the late nineties. Huge amounts of data can be automatically processed, thus relieving administrative staff from manual work. The free capacities could be used for in depth financial and economic analysis, human resource development, strategic purchase and so on. In fact, many once administrative departments claim to be strategic now ...

... the problem is that in many cases their staff was trained to do repetitive manual tasks in order to keep administrative processes standard and it seems not to be able to cope with the new "strategic" mission, consultants and training institutes have designed for them as an "unavoidable trend".

This paper will present the transformational work done at DaimlerChrysler in Spain, where the one-time conventional accounting area has become one of the drivers of the business by becoming a specialist in cost cutting via process optimization. Employee participation was key to liberate hidden knowledge and make invisible work valuable.

Present efficiencies often bear the seed of future rigidities and lack of adaptability. This is true for many business organizations that embark in restructuring and managers often end up wondering how their once efficient and well run organizations have lost their competitive capabilities. The solutions adopted often end up being layoffs. After expensive and dramatic re-engineering, organizations find several years later that they again have lost their momentum and the wheel of restructuring begins again.

How can organizations avoid this cycle of endless restructurings and job losses? Using the case of the accounting at DaimlerChrysler in Spain, this article shows how to identify organizational dynamics that produce deskilling and inefficiencies within business. It illustrates also how these processes can be reversed by involving employees in organizational development and change through Participatory Action Research (PAR), which will be explained later.


How Are Inefficiencies Born?

In order to explain the work done at DaimlerChrysler in making employee knowledge visible and putting it to work it is necessary to understand how deskilling processes work in corporations. As Braverman already pointed in 1974 this is a general dynamic in most companies.

Standardization and automation leave usually the staff doing more simple jobs in order to produce more output. But sooner or later new ways of production and new technologies will require expertise that does not exist in the company. The employees that were qualified for their old jobs are no longer qualified for the new tasks. Expensive recycling or layoffs are necessary as well as experts able to cope with new technologies and methods. These experts are generally difficult to find (until the new knowledge is standard) and therefore expensive to pay. Soon, we find a polarized labor market between the low qualified workers and the high paid experts. As time passes, the expert's work is also standardized, automated or off shored to low cost countries, leaving the former experts deskilled in relation to new developments.

Involuntary deskilling processes produced by companies

Figure 1: Involuntary deskilling processes produced by companies (click image for larger version)

Deskilled staff doing repetitive work seem unable to innovate and to "add value". Management sees them as having poor knowledge and related just to the few simple tasks they perform. Little knowledge about broader relationships between their tasks and other activities is expected from them. Soon, they are treated as "the" problem to be solved. Many companies solve the problem with big layoffs of those workers "that do not adapt themselves to changing environments". Another favorite measure is to export those jobs to low cost countries. In the short-term surely there is a gain. But in the mid-term it is doubtful that this strategy always pays off.


A Short Note on the Method Used

In this work two powerful instruments were used: participatory action research (PAR) and ethnography. Action research is a process where rigorous and systematic reflection of operational work in an organization is put in the form of an academic document. This document is then used as a key element of that operational work and, for scientific and methodological data gathering, is employed to nourish reflection. The output is both a scientific and theoretical contribution as well as operational problem solving for the client. Additionally, participatory action research means to involve the actors of the process in both parts of the outcome: the solution of operational problems and the writing of the scientific text.

In the end, the action research approach was chosen because the project was about solving a complex organizational problem for which conventional business approaches were not sufficient. Without a deeper understanding of the complex social and political dynamics operating within the organization, it would have been difficult to achieve the transformation demanded by management. The organization needed not only organizational transformation (processes, information systems), but also social transformation (people, working cultures, management style, knowledge, value). PAR allows the kind of intensive co-operation that results in organizational learning; something the organization dramatically lacked several years ago.

Action research in business organizations has a rich tradition that can be traced back to the Tavistock Institute. Especially since the early 1970s, companies have shown growing interest in this kind of methodology for problem solving. The pioneering work of Peter Checkland in developing the so-called soft systems methodology for information technology (IT) and information systems puts together problem solving (action), empirical work (research) and actors engagement (participation). Soft systems methodologies have not only been used in information systems design, but in industrial restructuring, performance measurement, process analysis and collaborative job design. Applications have grown over time, for instance, and are now being used in such complex projects as building networks of virtual enterprises (Löh, Katzy, Booth, Faughy and Thompson, 2003) or restructuring de Norwegian fish industry (Greenwood 1999).

PAR in business does not have a prescriptive methodology, but rather consists of a set of methods that place emphasis on practical problems in social sciences, with a commitment to the involvement of those being studied and who, in a positivist approach, would be considered "objects of the study." Thus, in PAR, the objects become research co-subjects.

Dulmanis et al. list five characteristics of PAR:

  1. Those being studied participate in the research process
  2. Research incorporates ordinary or popular (tacit) knowledge
  3. Research focuses on power with a goal to empowerment
  4. Research seeks to raise awareness
  5. Research is tied directly to action.

Ethnography has a specific value for Participatory Action Research processes. Research is a broad term that may include any scientific discipline or any multidisciplinary approach. But when complex organizational problems arise – as was the case at DaimlerChrysler in Spain – a social and cultural perspective is needed. Anthropology and ethnography offer the theoretical and operational tools for data gathering in real organizations, for the management of actual social processes and relations, and for engaging with social transformation in these real environments.

Many of these processes are beyond the control of the researcher due to their own and autonomous dynamics but the researcher can document them. Fieldwork and participant observation are powerful instruments for research within organizations; instruments that are more effective combined with the proper theoretical frame to describe and understand the observed phenomena.

If the causes beyond the symptoms remain hidden, if complexity is not mapped, the problems will reproduce themselves and grow worse. For this reason, ethnography, appropriately practiced has found its way into problem solving in business organization in areas that are highly problematic and especially complex.

For example, we can refer to traffic control. As the work of Richard Harper (2002) for the Xerox Research Center shows, ethnography has been a key tool in designing the very complex interactions between operators, computers, and pilots in an environment marked by heavy air traffic, information overload information, and high tension and stress derived from the responsibility for hundreds of lives in short time frames. IT Consultants usually ask users what they do and how they work in order to design systems. But air traffic presents so many exceptions to standard processes, so many unforeseeable cases may appear, and so many informal human contacts are necessary in order to deal with emergencies that it is more effective to see and experience what really happens in a control tower.

It is not by chance that ethnography is used to design the air traffic IT architectures because air traffic is a collaborative system. Ethnographers are often required where many (social) actors are asked to co-operate. Published examples include knowledge management solutions, like the one designed with Vidar Hepsø's aid, linking off and on shore facilities at the Norwegian oil company Statoil. The work of Bonnie Nardi in IT at Lucent focuses also on collaborative systems, a fact that also helped her to theorize about the invisible work in organizations, like the work of secretaries or telephone operators (Nardi and Engeström 1998). This kind of work, like the work of accountants described further in the case study below, for instance, is usually invisible because it is undervalued economically. It is not thought to be "productive" or "value adding" by the company – especially by those departments that are able to represent themselves as valuable. But often it is precisely within such invisible structures where the highly valued "practical know-how for everyday business" is stored (Nardi 1998).


Case Study: Reversing Deskilling of an "Aged" Staff in an Automotive Company

In the mid-90s, the Spanish subsidiary of Mercedes-Benz started the automation of its accounting unit, introducing SAP for its financial processes. Mercedes-Benz in Spain, a wholesale car and industrial vehicle distributor, belongs to a DaimlerChrysler Holding group that owns several other organizations in Spain: factories, financial services, insurance, retail distributors, etc. These firms all had their own administrations, human resource areas, controllers and so on.

Until that time, the accounting area had been organized as a classical functional unit: many simple tasks were performed in a simple manner by employees. Accounting information was assembled in a factory-like way into more complex units by the next hierarchical level. There were five management levels; from the simple data gatherer to the Finance Director. Due to the simplicity of work, the company never hired trained accountants, but just administrative staff. Only middle managers were true accountants.

Another quality of the accounting area was that it was considered by other areas and by itself as "unproductive", and as a "necessary evil" just needed to fulfill legal requirements. This fact had an effect on the staff's poor skill levels. This undervaluation would turn out to be very important in producing low performance in the area and was a key element for the change process, as will be shown later.

Since SAP automated most of the steps of traditional functional accounting the company decided to create a shared services center that would provide accounting services for a number of the companies the group owned in Spain, optimizing that way its accounting capacities.

The questions for the company were: how can we make our staff able to work with the speed and the quality that shared services demand and how can we make our administrative-minded staff take on a customer orientation?

Ethnographic Research

In order to solve these issues, the company turned the problem into a research question and designed an action program based on the findings done by management and some employees involved as co-researchers.

The ethnographic part of the work was structured to find out what the systemic causes for the qualification level and the employee performance were. Over a period of six months, workshops with workers and managers combined with participant observation of the everyday praxis of the work gave great insight into the organizational reality. The "reading", as suggested by Geertz (1997), of conflictual episodes, as well as interviews with employees, middle and top managers, consultants, employees from other areas and other ethnographic work allowed following organizational diagnosis:

Industrial Labor Relations

As a result of functional division of labor, classical industrial labor relations survived in the accounting area, while the rest of the company had already gone to more flexible ways of managing human resources. Industrial relations can be identified by practices that exchange job security for employee loyalty (the rotation level is below 5%).

Gender, Class, and Age

While women are a minority (under 10%) in the star areas of the company (sales and marketing), they are overrepresented in accounting (over 55%). Most of them are over 45 years old, have been working over 20 years in the accounting area and 60% do not have university degree (30% are only high school graduates). The majority of these women were hired as administrative staff, expecting that they would leave the company when they got married – a typical practice in Spain until the mid-60s. But the company's growth made clerical staff necessary, so that they remained in the organization.

Since accounting was considered of less value than sales and marketing in an industrial-relations framed company, these women, equally considered of less social value until the eighties, were hired for accounting.

Undervaluation of Non Commercial Functions

As has been already stated, accounting was typically a function considered as a "cost factor" a "necessary evil". Employees of these areas were labeled as "unproductive" and as "bean counters" (1). Budget, professional opportunities, and personal treatment were always inferior to those found in sales and marketing.

Deskilling Processes

While the general level of qualification has been rising in Spain continuously since the early 60s, and the percentage of workers with university degrees has grown in an exponential way (by 600% since 1965 according to Miguélez and Prieto 1999, 38-42), in the accounting area of the studied company, this has not been the case. The qualifications of people hired in the sixties and seventies did not rise significantly due to lack of investment in training. Until 1980, this was because training was not an issue for Spanish companies and later it was due to the described undervaluation of accounting: it was not considered necessary to train "unproductive" workers.

Nevertheless, working with employees gave me the opportunity to witness how well they solved problems as they arose in everyday work. Also, I was able to see how well-experienced employees eased tensions when conflicts with users of accounting services arose. They had a great know-how about accounting work, processes, and solutions for unforeseen dilemmas; but they were not able to name that knowledge and therefore to communicate it to management and to external consultants as a form of expertise. For management and consultants, "the worker had been doing the same thing for the past 20 years". But this "the same" was a long succession of cases, problems, and IT systems managed over the years: managers and consultants saw only the frozen picture, but not the dynamic story of 20 years of rich and intensive work and problem solving, including at the IT level.


There was evidence that skill issues and undervaluation of accounting were related and also that the undervaluation of accounting had to do with the generally industrialist regime that assigned value according to a factory production mentality ("productive" versus "non productive"). There was also evidence that this led to the valuation of people by class and gender. Thus, it seemed clear that further research had to focus on how value is produced in accounting, as value was a key category with which management operated.

For that reason, the next steps in the ethnographic research led to an analysis of the real economic value of accounting for the company and to the ways the symbolic and social value assigned to accounting resulted in a disadvantage in power relations within the DaimlerChrysler Group in Spain.

To find the real contribution of accounting to the actual economic result, the research design included the purchase of a car at a point of sale, so that the real process could be traced backwards through the system. The first outcome was that sales processes were not as linear as flow charts depicted them to be. They consisted of cycles with many interfaces to different organizational units, processes like purchasing, logistics, production ... and accounting. And they took place both simultaneously and diachronically.

Accounting (both finance and analytical accounting) was involved at the very point of sale to final customers (those supposedly to "add value" because the money inputs come from them) at many stages. Among many others: financial planning, operational planning, price building, calculations for commercial and marketing actions: discounts, gifts, etc., invoicing (communication between final customer, factory and finance), payment conditions, claims adjustments, cash flows, and information production for management. It turned out, that 53% of all the operations involved in every single sale originated in accounting. Other so-called "back office" activities (IT, Human Resources etc.) added up to 18% of operations. So, for every single sale, only 27% were purely commercial activities.


At this stage of the research we had sufficient elements to design an action plan together with management and the CO-researchers among the staff, as was described in the introduction to this case. It seemed insufficient merely to counteract the immediate effects that caused the situation or to treat the symptoms. We felt the action plan should provide a new strategic direction for the shared services in accounting according to their real value and avoid the repetition of the deskilling cycle. The diagnosis was that undervaluation of accounting functions had lead to hiring of under qualified employees, preventing the company from investing in training and also maintaining a industrialist human resource scheme for the accounting area.


Preview on Part II

Part II of the article presents the action plan derived from the case study reported in Part I. It reports the results and derives some lessons for practitioners and researchers from the experience.



  • Argyris, C., Putnam, R. And McLain Smith, D. (1985): Action Science. Jossey- Bass- Inc, Publishers, San Francisco.
  • Argyris, Chris (2001): Sobre el Aprendizaje Organizacional. Oxford, México.
  • Braverman, Harry (1998): Labor and Monopoly Capital. Monthly Review Press, New York.
  • Cernea, Michael (1995): Primero la gente. Variables Sociológicas en el Desarrollo Rural. Fondo de Cultura Económica, México.
  • Crump, Thomas (1993): La Antropología de los Números. Alianza, Madrid.
  • Dulmanis, McKenzie, Krooglik and Pejnovic (2004): Action Research and its application in business based organisational settings.
  • Geertz, Clifford (1997): La interpretación de las culturas. Gedisa, Barcelona.
  • Giroux, Gary (2001): Accounting history. Accounting History Page:
  • Greenwood, D. and Levin, M. (1998): Introduction to Action Research. Sage Publications, Thousand Oaks.
  • Greenwood, Davydd (editor) (1999): Action ResearchFrom practice to writing in an international action research development program. John Benjamins, Amsterdam.
  • Greenwood, Davyyd (2003): Interview. AIBR, nº 26,
  • Harper, Richard (2002): Rogues in the air: an ethnomethodology of 'conflict' in socially organized airspace. Xerox Research Center.
  • Harper, Richard (2002): The ethonoraphic turn: why it has come about and how to do it. Xerox Research Center., January 2004
  • Hepsø, V. Borstad, A.J., Midtlyng, J.O. (2002): CSCW-design and implementation compromises.
  • Hepsø, Vidar (1999): The development and coaching of Visok, a web based project in Statoil. ACM-Press, New York.
  • Holian, Rosalie (1999): Doing research in my own organisation: ethical dilemmas, hopes and triumphs. Action Research International.
  • Johnson, H.T, and Kaplan, R.S (1987): Relevance lost. The rise and fall of management accounting. Harvard Business School Press, Boston.
  • Lau Revil, Anne (2001): Shared objectives, shared understandings? Consequences of the global knowledeg industry. Unpublished thesis in partial fulfillment, Oslo.
  • Löh, Katzy, Booth, Faughy and Thompson (2003) Incubating Virtual Enterprise Networks in Yorkshire - an Action Research Approach.
  • Miguélez, F. y Prieto, C. (dir. & co-ord.) (1999): Las relaciones de Empleo en España. Siglo XXI, Madrid.
  • Nardi, Bonnie (2000): An Ecological Approach to Design. http://www.nardi/
  • Nardi, Bonnie and Engeström, Yrjö (1998): A web on the wind: the structure of invisible work. Computer Supported Cooperative Work, volume 8, numbers 1-2 (special issue) at
  • Pffefer, Jeffrey (1998): La ecuación humana. Gestión 2000, Barcelona.



  1. "Chupatintas" in original Spanish.


top top