Since early in the 1980ies, I had been eager to become self-employed. One of the reasons was that (as many founders I know today, too), I wanted to have a word and some responsibility in an enterprise and thus enjoy my work more. Another reason (and, surprisingly enough, not many of the founders I know today share this sentiment) was that I actually wanted to earn more money.
Consequently, I looked for the “ideal partner” (rather than the “ideal business idea”, because, even then, I thought there is no such thing). But it was not easy to even find the “ideal partner”. Luckily, though, I found Wolf Geldmacher after a little more than one year. He brought great entrepreneurial strength and was just as “down to earth” as I was.
With Wolf as my partner, it did not take long to establish the “InterFace Connection Gesellschaft für Datenfernverarbeitung und Entwicklung von Software mbH“, which was the predecessor of InterFace AG. Our fields of interest were IT and Unix. We wanted to build a successful product with the “new” Unix. We assumed that services are hard to scale. Even before 1983 (the founding was in 1984), we feared that Body-Leasing might not be a business with a long life-span. Besides, even in those early days, it seemed to us that the Body Leasing Business is situated somewhere in the legal grey zone. This was one of the reasons why we wanted to actually create a real product.
Said product was the office-compatible Unix writing system HIT. In retrospect, the entire affair was a totally crazy enterprise. Surprisingly, though, it worked. I fact, to this day I cannot entirely understand how this happened.
It only took as a few years to become the most successful Unix text system in Europe by far. It felt like a dream!
Looking back, I can identify persons and important requirements or events without which it would never have worked. Basically, we were moving at the right time and had lots of luck, because everything just fell into place.
The Duo „Wolf & Roland“
Even in the early 1980ies, we believed in “agile, lean and open“. We were in favour of self-organization and self-determination, We communicated our ideas and our expectations and then our teams were allowed to work as they saw fit. It was all done with a huge sense of belonging together.
For software development, we invented and practiced a private method you would, today, call SCRUM. Wolf was the “SCRUM-Master“ (and more than that, too). He was responsible for the technology and the people. He led the colleagues towards delivering quality and made it clear to them that they had to live quality – primarily for their own sake. And I was something like the “product owner”, the leader on the business side.
In 1960, my parents sent me to the business-centred Higher Middle School “Jakob Fugger” in Augsburg. Later, it was re-named Grammar School, before 1960, it had been called business school. Stenography and typing had been compulsory until 1960, later they were taught on a voluntary basis. My father forced me to learn both, because he considered both competences as indispensable when it came to the real fight in professional life.
Anton Böck was my teacher. I was doing quite excellent in stenography. Whenever I was forced to do some learning at home, I spent hours drawing stenographic symbols. For me, this was just like calligraphy, simply beautiful. I hated the typewriter. Herr Böck was a strict teacher and he rather liked me because of my stenography. Consequently, he forced me to practice on the typewriter. As a result, I dreamed what characteristics a “nice” and “lovely” typewriter should have as early as by the age of 16.
Well, it might sound ridiculous, but I am quite sure that, without this early experience with generating text, the InterFace Connection would not have become a product enterprise.
Hans was my mentor and the man who made UNIX big, both in Europe and at Siemens (where it was called Sinix). I was convinced that all he said was correct and he believed in our team. Well, this facilitated matters a lot.
However, our success proves that we were correct.
Dr. Peter Schnupp
Peter was an IT pioneer of the second generation (for me, the generation Zuse was the first generation and I was part of the third). As an entrepreneur (he founded Softlab), IT expert, column author in the “Computer-Woche” and also due to other activities, we was well-known and had an excellent reputation as an expert.
Peter managed to convince the strategic decider in a big public office organization that the future will be based on IT and UNIX – and that there is a great local product.
Without this stroke of luck, the project CLOU/HIT would never have become a success.
Even as a young software developer at Siemens AG in the mid-seventies, I had a great job. In the course of developing Transdata, I designed the “Connection Handling” and also worked on the development of “APS“ (appliers’ programming language). Connection handling has a central function in “long-distance data transfer”, as it was then called. Through APS, you could outsource processing time to local “data station computers” (operating system PDN) – which was the first time the central principle of main frames was broken.
With this “superior knowledge”, it was easy for me to get a good reputation with huge projects, which meant that, as a logical conclusion, I was transferred to the Siemens AG department: “special projects marketing”. My most important project in this department was DISPOL, a central project of the Bavarian Police which had set itself the task of making the filing cabinet (data), the typewriter (documents) and the telefax machine (communication) obsolete through the introduction of Electronic Data Processing.
I stayed with this project until I founded my own company and in the process I learned all there is to learn about the expectations of customers and the market of civil offices.
Without this history, HIT/CLOU would never have become a successful product.
For product development, we hired very young persons. In fact, they often started working for us when they were still students. And in (almost) all cases, they were exactly the right people. These persons took no time at all becoming central top performers and they instantly took high responsibility.
The Right Principles
Along with the product development, a qualified counselling team in the field of “Unix Operating Systems” grew. We sat right at the source and I learned many extremely helpful things from our operating systems partner. Consequently, we were early to use tools which at the time were not even widely known in Europe.
We applied methods (or rather, we intuitively invented them) such as the four-eyes-principle when programming, peer2peer-reviews, ”extreme programming“, developer rotation and much more. Those were methods that had not even existed in those days, yet they gave us considerable advantage when it came to developing speed, customer orientation and quality (incidentally, some of these principles are still not applied everywhere today).
Our developers are always directly in contact with the customer. For instance, all the developers did the HIT courses for the customers themselves, which means that they understand what the customer wants and needs.
All these factors have contributed hugely towards the quality of our product.
We were well aware what the real problem with our project was. Consequently, we shared the costs of the IF computers during the first phase of basic development. For the second phase of marketing, we had intended to share tasks. We were going to further develop the basic system and also attend to our expected big customer Siemens. InterFace Computer was planned to be the port for many other Unix systems and for realizing the sales for other partners.
But even in order to just develop the product, we needed considerable man power. In 1984, we solved this problem in a very simple way: Wolf Geldmacher and yours truly worked as counsellors. We took care of the product and team in the evenings and, when necessary, on Saturdays.
As counsellors, we worked for an hourly rate of 150.- DM. This was absolutely top range and we were only able to get away with it because US consultants with comparable know-how were even more expensive.
Well, let me do the calculations: a good month will bring us 200 man hours (we were very industrious). Multiplied by 150 DM, that equals 30,000 DM in a good month. With a salary of 5,000 DM, which is about 6,000 DM before taxes, that left us 18,000 for hardware, our Heidi (our assistant, who helped us from the beginning) and our students, the product developers.
Only a few months after our foundation on April, 1st, 1984, we were also able to recruit two young computer scientists who immediately worked as consultants, too. They brought in money comparable to our ratio. And from late in 1995, the product itself provided us with rapidly increasing amounts of money.
There were more beneficial circumstances that helped us considerably.
For instance, Siemens had started a very huge project the aim of which it was to develop their own text system for BS 2000 and Unix. Regardless of the fact that this project had manpower that was several times more than our development team, on top of all Siemens project developers being full-fledged software developers (as opposed to our young team), these projects never seemed to get anywhere. Which means they actually more or less failed.
The Siemens concern, however, needed this software for their goals and therefore had to buy licences from two suppliers – one of which was us. This is how we became the supplier of the then market leader in Germany of Unix.
The technological time frame also favoured us. Firstly, Unix at the time replaced all those many different computing systems of the “medium data technology” MDT. So again, our product HIT came just at the right time.
Also, this was the time when using databases started to get a common thing to do. SQL as “query language” based on natural language was newly defined. There was even a German version of SQL!
So wasn’t it only logical to extend the 4GL CLOU (for programming text elements) which supplemented HIT by an “embedded SQL“ which would suddenly make it possible to send dynamically generated queries to a database during the running of the text-element program and then to automatically make use of the found data for producing a document? This was a true sensation, and it had come at exactly the right time.
Lots of Luck and Just a Little Bit of Misfortune
The courage shown by a big federal office in relying on a totally new technology and a very small provider was certainly something special. It was a wonderful market development in favour of Unix. There were many more courageous and, for us, beneficial decisions by customers. A great team…
We Also Had Some Problems
The hardware we needed for our developments was extremely costly. As early as 1985, we had to buy an MX500 – which at the time had a listed price of several 100,000 DM. For us, this was impossibly much money. But it was absolutely clear that we were not going to manage the necessary developing speed without the system. And then it took only two years for the machine to become outdated. Overnight, we developed on SUN and on the new fast PC-s with diverse Unix variants.
In the long run, it turned out that InterFace Computer, too, was the wrong partner. The strategic cooperation no longer worked. Consequently, we were forced to buy the rights to the product. This was a truly hard investment decision. Yet, in retrospect, it turns out that it paid off.
Other Necessary Requirements
I am sure there were other causes and lucky circumstances without which the enterprise HIT/CLOU would never have made it. Partly, we are probably talking things I no longer remember or things I was never really aware of. But without all this, at least InterFace Connection as producer of HIT/CLOU would never have existed. If you remember what I wrote, you will see that I kept telling you at every possible instance that it would never have happened without those circumstances/ coincidences. They were all necessary for the final success. And in retrospect, I am truly surprised how this courageous enterprise could actually work out.
This article was supposed to show you through my own example how unbelievably many requirements have to be fulfilled in order to become a success. Mind you, I do not at all want this to discourage you. But it certainly goes to show that it is not all that easy and that a pragmatic approach might sometimes do no harm. I believe founders can learn a lot from this story and I am quite willing to discuss it with you as a Use Case interactively and personally.
(Translated by EG)