I frequently come across tenders whose aim is to standardise an existing, historically grown service. Until now, this service has been provided in the most diverse variations and qualities by a wide variety of people. After several company mergers, local IT in the countries, different cost centres and various internal customers, things can get quite confusing.
Many consultants recommend in such a situation that you only have to describe exactly what kind of service you need and demand corresponding service level targets from the providers. In many cases, this is an almost impossible task. The very fact that the organisation has to cope with different models in different business units. Makes it difficult to define the right service.
Such tenders are usually very obvious. When reviewing them, it is noticeable that the numbers do not match and that the required service levels are usually too high and hence too expensive. In further communication, the figures on the client side are typically corrected every few weeks, but remain unsatisfactory until the end.
I myself worked on the client side for several years and reluctantly recall the simplifications of central stakeholders with the blunt statement: “What’s so difficult about that?” I would like to give an answer to this. A structure that has grown over years without standardisation with hundreds of different user groups is complex, perhaps even chaotic. It is more than complicated and cannot be analysed adequately in a short time.
With the help of Cynefin, I would analyse the environment and instead of a simple answer, I would take a complex problem, for example, and respond with Probing -> Sensing -> Acting (Emerging Practice). What does this mean for a tender? For example, choose an adaptive tendering method or, if you want to keep it traditional, learn and standardise in phases using a framework contract.
If you start from the Cynefin categories and choose a simpler context than is actually given, you have fewer steering means than needed. Ashby’s law states that the variety of the control system must be at least as large as the variety of perturbations that occur in order to carry out control. In other words, category errors are high risk, intentional category errors are gambling.
And that is precisely my observation in such circumstances.
