Artur Södler Software-Qualität

What is to be done after all?

Operating results are effected by company policy. That is true for profit or loss as well as quality of products. If the shoe is already pinching, the management must eventually ask itself the unpleasant question whether it did set the course poorly in the past. Now we must not allow investigation of causes and looking for scapegoats. Instead, it is all about preparing the ground for better quality in order to let the staff do the job.

Both individual responsibility and cooperation are needed. The company policy should enforce that everybody is responsible for his own work. Everybody must plan necessary resources in advance, ascertain them, and — also in advance — negotiate the expected result with the recipient of the service. If expectations are set unilaterally, or if required resources are withdrawn, then this must be considered equivalent to taking responsibility.

Planning is part of the work and a hierarchic task. The VW management board decides on a new model series and its overall concept, but not which tin plates, cables and control devices are to be built in. A separation of project management and software development as may be the case is to be scrutinized, for software development requires at least leading subprojects.

Quality requires investment in training. But as long as results cannot be quantified, companies shy away from that investment. Hence the first step is to enable the company management to judge such results and to make them comparable to other figures of the company.
Not till then the company management can point the need and the value of such an investment to the employees so that they consider the time for training and research as an inherent part of their project planning, no matter on which level of detail.

Teachers and books transfer knowledge. Indeed, we learn most intensely from errors, but only what is to be avoided. How to do things right is taught by a vast amount of books and internet articles. Educationists bridge the gap, find the adequate teaching material, and motivate to learn.

Look beyond your own horizon. We are all but small ants in a large universe. Just as small is our knowledge. People who dare to look beyond their horizon can forebode what is beyond and prepare themselves. If developers expect at the outset a software can be used on other conditions, then it is not much effort to change the software into an all-purpose tool.

People do not think faster under pressure. (From: Tom DeMarco, "The Deadline", ISBN 0932633390) Their brains work faster, but less qualified. First deduction: Deadline pressure must be avoided, best while meeting all deadlines. Second deduction: Sugarcoated deadline commitments are off-limits for the marketing department to reel in cusomers. Commit down-to-earth dates to your customers, which you can guarantee, and keep your customers up-to-date as soon as it becomes apparent that the job is completed earlier. Agree upon smaller projects, make a 90% success out of a 50% effort. Then see what can be done next.

Planning and software development are creative work. The easiest way to suppress creativity is by doorstoppers named "guidelines" on how to create software. Extremely thin guidelines can help to speak a common language in order to make work results exchangable and meshing.
Furthermore, new ideas must never be dropped arbitrarily or for no reason at all. Nobody knows what is being invented next: The slider bar, Dadaism, multitasking, interactive arts — what will be next? Indeed many ideas turn out to be a flop, but if we use this as an occasion to nip creativity in the bud, we will never find any genious solution.

Create a positive atmosphere. That does not mean flower-boxes! Computer nerds can't do anything with it anyway. You have studied business economics and you know "The Peter Principle" (Raymond Hull, Laurence J. Peter, ISBN 0688275443)? Well. But business management is also a creative job, to which dissenting ideas are of importance. Why don't you read the chapter "New Company Model: OA5" from the book "The Dilbert Principle" (Scott Adams, ISBN 0752272209)?

There's nothing good unless you do it. Go for it.

We hope you already found some good idea here. We are even more pleased if you worked up some curiosity. Please remember: There's no harm in asking. So best talk to us directly, for the personal conversation is best to show whether and how we can help you.

Artur Södler Software Qualität