Business Process Transformation - Spring 1996 Business Process Transformation - Spring 1996
In this page we consider three different ways of viewing processes at the macro level. True innovation (rather than improvement) is more likely to be achieved at this level.
Davenport [chaps 11-13] classifies processes within a typical manufacturing organisation and outlines process innovation strategies for each category. This may give us inspiration for our process visions. Of course in a service organisation (such as a university) some classifications will differ but the examples will still be helpful.
These processes include research, design, manufacturing and logistics.
Research is seldom viewed as a process (in the BPR sense) but in both business and government research is usually directed to some practical end. Thus, while it might be difficult to predict the exact outcome or duration of a particular research project, a focus on outcomes can be the basis for process innovation. Particular improvements might include better coordination amongst researchers and improved access to external sources of information (eg use of the internet).
Design is also difficult to organise, but is clearly outcome directed. Design processes can be improved by designing various parts of a product in parallel. To achieve this often requires networked computer aided design (CAD) tools so that different design teams can easily integrate their work. Other innovations might aim to increase reuse of existing components (with an expert component librarian) and to streamline consultation with customers. Design decisions also need to be made quickly.
Manufacture, in the traditional sense, is the area in which process innovation is already most advanced. Quality adding processes, inventory management, just-in-time production lines and manufacture to order are common approaches to process improvement. In service industries it is sometimes difficult to find the equivalent of "manufacturing" processes. Activities where the customers are directly involved, such as nursing, teaching or parts of banking are better considered under "customer facing processes" in the next section. Activities such as repairing products, preparing teaching materials or drawing up contracts are however akin to manufacturing, and similar process visions might be helpful.
Logistics involves getting the right things in the right place at the right time. This can range from getting purchased products to customers to getting students into tutorials. Logistics is another area where a lot of reengineering has already been done (though not necessarily with the UTS timetable). Common principles seem to be: concentrate on the process, rather than trying to optimise resource use (ie warehouses, classrooms); schedule preparatory sub-processes (eg credit checks, subject offerings) early; make information available where and when it's needed.
In a sense (as long as we hold a BPR philosophy) all processes are or should be customer facing. Even raw materials acquisition has as its ultimate end the delivery of the product to the customer (and the receipt of the customer's payment). Here we deal with processes which involve direct dealing with customers. These can be classified as marketing, sales and service.
"Marketing-oriented processes increase the likelihood that a customer will engage in a transaction or a relationship." [Davenport, p244] Because the result is probabalistic it is difficult to define firm process outcomes. However there are many subsidiary processes which are candidates for reengineering - knowing what is for sale, making contacts, producing advertising campaigns, even collecting statistics can all be improved. (I think Davenport's examples are poor here - he concentrates on data collection and modelling, which are not necessarily process oriented.)
The sales process runs from the time the customer expresses an intention to purchase the product to the satisfactory delivery of the product and payment by the customer (note that this includes some processes included under logistics previously - depending on how broad a view we take of processes). Activities include checking price and availability, credit checking, initiating manufacture (if made to order), delivery and billing. Unless the sale is seen as a single process it is easy for the sale to be seriously delayed at functional boundaries. There are many opportunities for innovation in all parts of this process.
Service is similar to sales but demand is often sudden and unpredictable, and customers are often anxious for quick turn around. One striking form of innovation is remote servicing, where a device may call the service centre when it "feels sick", and technicians may be able to fix the device by remote control (especially if it is a software problem).
Case Management is a popular reengineering strategy for both sales and service. This means that one person (or team) is responsible for pushing through the whole process, representing the customer within the organisation. This person needs good access to all data, people and activities related to the process. IT support is often critical for successful case management.
A variation on case management is management by the customer. Here the customer has sufficient information and influence to follow their own interests within the organisation. This is sometimes done by very powerful customers (for example, the government may place agents in the factories of defence contractors) but with the advent of electronic commerce this may become possible for individual customers.
Besides the fact that managers resist being redesigned, the problem with applying process innovation to management is that it is often not clear what the processes are meant to do. Many mangement processes are supporting rather than strategic. Of course particular processes such as having meetings or producing strategic plans can be improved but this may not have an obvious business benefit. For major gains we first need to determine what management processes should be, using some other technique such as soft systems methodology or the viable system model.
In the section on case management I mentioned customer managed processes. Most critical business processes involve relationships with other organisations. Some of the most successful process innovations have involved altering organisation boundaries, move some activities from one organisation to another. Concepts to help us with this type of innovation are discussed in the following sections on value chains and transaction cost modelling.
In his well known book Competitive Advantage Michael E. Porter takes what is essentially a process view. He looks at the main business of an enterprise as a single process and divides it into a number of subprocesses. He then looks at how each subprocess adds value to the overall process. Suppose a restaurant turns raw materials into food delivered to satisfied customers. We can cost the raw materials and we know how much is paid for the final dishes, so we know how much value has been added in total. In principle we can now divide this added value among the activities such as purchasing materials, taking orders, preparing ingredients, cooking, presentation, delivery to table and collecting payment. Peripheral activities such as decorating the room also add value, since people are prepared to pay more for the same food served in a better atmosphere. Having assigned values to the activities we now have a "value chain".
If the value chain can be established it gives us a number of clues for reengineering. If complex or expensive activities are adding little value we may wonder what use they are. Ohterwise, when deciding which processes are the most suitable candidates for reengineering we would look to those which add the most value because there we would expect the greatest gains.
To help establish value chains in complex organisations, Porter has defined a number of typical key activities.
inbound logistics, operations, outbound logistics, marketing & sales, service
procurement, technolology development, human resource
management,
maintaining infrastructure for planning, accounting,
finance, legal matters, government liaison, quality
The first difficulty comes in identifying the main business. The metaphor here is the "powerhouse" of the enterprise. For car manufacturers (say) this might seem obvious, but there may be disagreement as to whether design or production is the main business. IBM from time to time claims to be in the business of marketing! For institutions such as banks the main business is difficult to identify. Some government enterprises seem to be mainly in the business of (literally) selling themselves.
Secondly, it is very difficult to discover the value added? How to compare the value of an exquisite Hollandaise sauce with soft lights and sweet music. The only sure way is by having a competitive market at each stage. Then we can deduce that the value of cutting potatoes is the difference between the price of whole potatoes and pre-cut chips, that the value of the sauce is the extra price we can charge compared to the nearby restaurant with the same music. This can all be quite difficult. (We can see why economists like to have a free market in everything.) So in fact, "value added" is often based on cost. That is we assume that the value added by each activity is proportional to the cost of performing that activity. From the point of view of reengineering, this prevents us from identifying high cost, low value activities, but since our other guideline is to prioritise "high value" activities the former may come to our attention anyway.
MacDonald [in Scott Morton, Appendix D] proposes an extension of the value chain called the Value Process Model. This adds to the value chain a more detailed (diagrammatic) description of work processes with fairly detailed costing (as a surrogate for value analysis). MacDonald claims that this shows us what is actually happening, rather than concentrating on outcomes.
MacDonald says that process costs fall into 5 major categories:
investment, input, process, anticipatory and fixing up
costs,
overheads
He also discusses suitable change criteria, concluding that usual criteria such as return on investment are too difficult to calculate and too closely linked to existing structures. His recommended criteria are: lower costs, process effectiveness, ouput quality, business expansion or change. MacDonald's approach is closer to conventional business redesign theories but still allows the possibility of radical change.
Transaction costs are discussed in a number of theories, including transaction cost economics and transaction cost analysis. The general idea is to take account of the costs involved in relationships between organisations. For instance, the savings gained by obtaining a product at the best price may be outweighed by the cost of finding the cheapest supplier. This leads to questions of alliances, relationships and trust. (How much is it worth to trust your supplier to always give you the best price? Is this trustworthy behaviour backed up by self-interest?) Transaction cost economics tries to model the effects of relationship costs on diecision making. An example of how these issues can be included in process modelling is Scherr's use of speech act theory.
Transaction cost analysis is a little different, in that it looks at the structure of interorganisational transactions. These structures relate to J. D. Thompson's categories of interdependence. Pooled interdependence describes a situation where "partners" need to cooperate at the minimum level, with both sides acting totally according to their own obvious self-interest. The structure for this type of interaction is the market. Sequential interdependence means that one partner must complete an activity before the other partner can proceed. This leads to an hierarchical structure, where control is exercised by one of the partners, a third party, or an agreed set of rules. Reciprocal interdependence means that each partner is dependent on the other's actions. Here the partners will form a "clan", based on mutual respect and (dis)trust.
Transaction cost analysis states that costs of transactions increase with the complexity of interaction. That is, markets are the least expensive to maintain, clans the most. This has two implications for process reengineering. Firstly, since reciprocal interdependence is the most costly, redesign in this area promises the greatest benefits. Secondly, if organisational boundaries could be moved so that more complex transactions can be replaced by market transactions then great improvements could be achieved.
A not very exact example: Suppose we wish to support users of a particular piece of software. The reciprocal interdependence model would provide real-time help in person, by telephone or by internet "chat". (To really fit the clan model, the support person would have to develop a personal relationship with the customer - as with an on site engineer.) The sequential model might be via e-mail, fax or post. An enquiry would be sent in, and an answer received later. With the pooled model we might supply the user with a manual or on line help which they could access as the need arose. The latter is obviously the cheapest solution for more than a small number of queries. If we could design it to provide customer satisfaction it would be the best.
BPT Lecture 6, 13/9/96.
Any questions or comments, just e-mail me.
This page is maintained by
Jim Underwood
who can be reached at
jim@socs.uts.edu.au.
This page was last updated on September 12th, 1996.
http://linus.socs.uts.edu.au/~jim/bpt/procbig.html
