As already indicated, the engineer tries to analyze everything precisely, plan and execute it to perfection. The business manager, on the contrary, first separates the more important aspects from the unimportant, and then concentrates his efforts only on the important ones. The procedure known as the ABC-Analysis turns this into a systematic process.

ABC – basic concepts

The basic idea of the ABC-Analysis comes from the observation that, regardless of the domain, it is almost always a few individual components that determine a large part of the costs. For example, look at a car: car body, engine and gears are the key cost factors here. Although these are just three items, they determine about two thirds of the production costs. These parts are referred to as the "A parts". In practice, it is usually about ten percent of all parts that determine about 70 percent of the costs.

The next group consists of parts which still account for a substantial, but considerably smaller share of costs: for example, brake units, air conditioners or central locking systems. Even these so-called "B parts" still have a tangible, though significantly lesser impact on the total costs. In practice, around 20 percent of all components can be classified as B parts. Together with the A parts, these already account for nearly 90 of the total costs!

The remaining majority of the components, namely about 70 percent, hardly have any perceptible impact on the costs. These are the so-called "C parts".

This is the characteristic curve obtained when this is represented graphically.

Fig. 1: Classical ABC-Analysis

Now, if the manufacturer is successful in reducing the manufacturing costs by just one percent in the production of the car body (A part), on the whole, this amount will be far more effective than, say, halving the costs for the windshield wiper (C part). All further efforts should therefore concentrate especially on A parts, and, with some reservations, on B parts. Savings in the C parts are comparatively uninteresting.

Application to technical documentation

The basic concept of the ABC-Analysis can be applied very well to the field of technical documentation. Let us consider, for example, the individual topics in a manual, or an online help, as the individual parts or components of the product called "documentation". If we examine how frequently users consult these topics, we again get the familiar curve.

Fig. 2: ABC-Analysis of topics in documentation

About ten percent of all topics take the major share of about 70 percent of all accesses (A topics), while the large majority of topics remains virtually unread. So can you really justify to describe all topics by putting in the same kind of effort? We will address this question once again later on, while discussing quality planning.

The ABC-Analysis can also be applied beautifully to processes and project plans as well. Almost always, it is only a few steps that cause the lion’s share of problems. We should invest all our energy in optimizing these few steps, and ignore the rest. We will come back later to this issue as well.

In the first step we got to know an elementary procedure for separating the grain from the chaff, through the ABC-Analysis. The second basic procedure that you will need to master for cost controlling is the use value analysis. A use value analysis will help you in making (investment) decisions systematically and objectively. The main advantages compared to decisions that are based on "gut feeling" are evident:

A sample analysis

How a use value analysis is done is best illustrated by an example. Imagine you are looking for a new authoring tool to create manuals and online help for a new line of products from a single source. The documentation is expected to run into 1,000 to 2,000 pages on the whole, and would also need to be translated into Chinese. Since the scope of the documentation is fairly modest, you have already ruled out expensive content management systems to begin with, as well as all programs which are not at all capable of single sourcing, or which do not support Asian languages. To keep the effort and money spent on the use value analysis within reasonable limits, it is useful to concentrate exclusively on those alternatives which actually fit into your narrowed down selection. It is only in exceptional cases that you will encounter more than four or five alternatives.

In our example, the following three investment options remain after your selection: "RotoHelp", "AuthorMe" and "FrameTaker + WebWordsPublisher".

In the first step of the use value analysis, you will decide on the criteria that will influence your decision, and the weightage that you will assign to each of these criteria. It is not necessary that all criteria are quantifiable in figures. You can also work with purely qualitative criteria. In our example, a simple catalogue of criteria would appear as follows:

Make sure that the same criterion does not appear many times over in your list, under different names. This would distort the results due to multiple weightage. In more extensive use value analyses, with a larger number of criteria, related criteria are often grouped.

In the second step of the use value analysis, you will now evaluate the available alternatives with respect to the criteria listed. You will not need any costly and special software for this, a pocket calculator or any basic spreadsheet software will be totally adequate.

For making the evaluation, allot points from one to ten. One means that the criterion is not very adequately satisfied, ten means that the criterion is fulfilled optimally. Note that a high value of a certain criterion does not automatically correlate with a high rating. High costs for instance are undesirable, hence they will be assigned a low number of points.

In our example, the allocation of points could appear as follows:

Table 1: Use value analysis example

In the first column we have the criteria that were listed out earlier, in the second column we have their weightage. Let us see how the use values are calculated now, for the investment option "RotoHelp": RotoHelp gives very good quality online help, hence it receives the comparatively high rating of eight points. This means that the criteria "quality of the results online" gets a use value of 20 % x 8 = 1.6. Similarly, calculate the individual use values of all the other criteria. The total use value of the alternative "RotoHelp" is then calculated as the sum of all the individual use values.

In a similar way, evaluate the alternatives "AuthorMe" and "FrameTaker + WebWordsPublisher". The alternative that you should decide on is the one that has the greatest use value for you.

Important: The result of the use value analysis is your own, highly specific result, based on your individual criteria, weightages and evaluations, and suited to your particular situation. It is almost certain that if the decision were to be made by another person, the result would be very different. Similarly, you might perhaps have evaluated specific points quite differently from the way we have done in our example. This does not point to a weakness in the system. Rather, the strength of the use value analysis lies precisely in its ability to provide specific answers to different situations.

Sensitivity analysis

The results in our example are quite close. If this is the case, or if you are not entirely sure in individual evaluations, you should follow it up with a sensitivity analysis, also called a "What if analysis". Here, you can vary selected individual weightages and evaluations and observe the effect of the changes on the result. The more stable a result is at the top, the more reason you should opt for the particular alternative.

Taking costs into consideration

Almost all decisions in the business domain are associated with costs. The simplest thing to do, as we have done in our example here, is to include the costs directly as a criterion in the use value analysis. But if you want to compare costs and benefits explicitly, you can also first conduct the use value analysis without taking the costs into account. Then you can divide the use values by the corresponding costs to get the relative benefit per unit of money invested.

Deciding in a team

Important decisions are rarely taken alone, but rather by a team. This not only has the advantage that all members in the team can bring their understanding of the situation to bear on the decision: more important, anybody who is involved in a decision personally will later also support the implementation and not sabotage the decision, as it might be the case if he or she could not participate in the process.

You have three alternatives for doing this:

Alternative 1: Depending on the expert knowledge, each member of the team performs a particular part of the evaluations. This alternative is useful especially in situations where not all participants are equipped to make well-founded statements for all criteria.

Alternative 2: After a discussion, a group decides on a common evaluation on each point. The disadvantage in this alternative is that team members who are particularly strong in group discussions will dominate the decision.

Alternative 3: Every member of the team performs all the evaluations, alone. Then, the mean is calculated from the individual evaluations and the use value determined from this. This alternative is always recommended if, unlike in Alternative 1, no special expert knowledge is necessary for conducting the evaluation, or if all decision makers have the same expert knowledge.

Some more examples

The use value analysis is not restricted to the choice between investment alternatives. You can always employ the use value analysis whenever a cost consideration alone is not enough, as in the following cases:

In your day-to-day work you will surely come across many more cases where this method can be applied.

What defines the "quality" of documentation very much depends on the point of view. For an engineer, quality means first of all the document is free of errors and complete. For a controller, however, quality means minimum costs in terms of documentation cost on the one hand and support cost on the other.

So, one thing is evident: quality is something that needs to be discussed as early as possible. It is a lot cheaper to avoid errors than it is to correct them. The more advanced the project, the costlier the changes.

Fig. 3: Cost of changes as a function of project progress

When you agree upon concrete quality goals, preferably in writing, you can be relieved of the personal dilemma of never being able to achieve what you would like to achieve from an engineer’s point of view, due to cost and time restrictions. Always bear in mind that beyond a certain point, the costs for further improvement increase rapidly. It is never possible to achieve 100 percent with a reasonable amount of effort. During the quality planning, you should decide how far to the right of the curve you want to move. If you achieve the agreed quality at the planned costs, then, you will have fully achieved your project goal.

Fig. 4: Cost as a function of quality requirements

Expenses for fulfilling the functional requirements

As part of the quality planning, always ask whether the effort needed to fulfill a certain function is reasonable:

Utility or image

One useful aid in deciding which functions can be omitted is to differentiate between utility functions and prestige or image functions. A welding robot in an automobile manufacturing unit, for instance, has an almost exclusively usage oriented function, whereas many of the mobile phones used today have more to do with an image function.

Fig. 5: Utility and image

During the quality planning, sit together and decide where your product and your documentation will be located on the graphics. Depending on this, determine which features are really useful in your documentation or online help.

ABC quality planning

As explained already in the section on ABC-Analysis, it is possible to classify a chapter in a manual, or the topics in an online help as "A", "B" and "C". Topics which are most important, and therefore classified as "A topics" are

Typical C-topics no doubt outnumber all other topics, but are rarely read. Special reference topics are a typical example.

The effort, and with it, the costs for creating the individual topics can be classified as follows:

A-topics

B-topics

C-topics

The moment a project runs into delays, unplanned additional effort will be consumed for planning, crisis management and eventually also for involving more hands. Not to mention the loss in revenue if the product shipment is delayed because the documentation was not ready on time.

A, B and C tasks

As in the case of any production process, even in the "production" of technical documentation it is the bottlenecks, and these alone, that determine when the product will be ready. Any time that is lost in dealing with the bottleneck means time lost over the entire project, for ever. Even in the individual "manufacturing steps", we can once again differentiate between the "A", "B", and "C" tasks. Tasks related to dealing with bottlenecks are "A tasks", tasks which are not critical in time are the "C tasks". Special attention should be paid to the A tasks during the planning and execution. C tasks can be delegated and even disregarded.

Some typical examples of bottlenecks in the field of documentation are:

Measures to adopt for dealing with bottlenecks

Here are some of the most important measures for handling bottlenecks:

Try to change your perspective now and then, and look at your work from the point of view of a business manager:

Besides reducing the costs, a lean authoring process also brings in other improvements which are at least as valuable, if not more:

And finally, a stable process is also one of the basic pre-requisites for constant quality. That is why quality need not necessarily suffer when the costs sink.