Prioritizing work is quite the topic. Most of the time, work first gets prioritized when it is already in the system and, quite frequently, that leads to perpetual reprioritization. The criteria for reprioritization usually remain unclear, and they often only deal with the fear of possible consequences. However, assuming the priority is already classified with a logical approach before the work enters the system: Which approach should be used for it?
An economical approach makes the most sense. What I hear in the stories and questions of participants in my trainings is that, in many companies, there are indeed the beginnings of an economic approach to work prioritization. These approaches, however, are often not completely thought out. Value, or “business value”, is the prioritization tool used in such cases. It’s a great idea, but leads me immediately to the question: “How are you using it?” The answer to this question is usually: “Work with the highest value comes first, and those with lower value are implemented later.” Sounds logical, doesn’t it? Yes, as long as you don’t consider the amount of time needed to implement the work. When work X generates high value, and needs four weeks to be completed, while work Y can generate the same value in two weeks, which one gets implemented first? It probably makes more sense to complete work Y first because, stated simply, it brings money in faster than X. When we use value as a prioritization attribute, it should always be considered in relation to the implementation time of the individual work.
Let’s go one step further: Is value a static quantity? Chocolate Santa Clauses do not sell well at Easter—in other words, the value of work can change over time. Or it can, subject to seasonal influences, only reach its full-potential at a specific time or within a specific timeframe. The change in value over time should absolutely be taken into account in a prioritization process. Now we have all the ingredients for prioritization based on Cost of Delay:

• How high is the value of the individual work?
• How long does it take to implement the individual work?
• How does the value of the work change over time?

Cost of Delay are those costs, as well as economic effects over time, which occur when the completion of work is delayed or doesn’t respond quickly enough to the market. Cost of Delay includes not only actual costs incurred, but also all lost revenues which are incurred, regardless if a project is worked on or not.  To be able to work on orders in an economically useful sequence, the Cost of Delay is quantified by visualizing it in concrete monetary amounts. Cost of Delay is always drawn from the value-generating elements (deliverable units), that give the customer a concrete benefit. It is a function of the value generated by an activity, and its urgency.
I recommend initially using Cost of Delay when the Kanban system is being designed with deeper understanding. What do I mean by this? The people assessing the individual implementation options allow other criteria to be used in their valuations, since there are several risks to consider. However, as long as you do not have other information or measurements available in the Kanban system, it is always better to work with the cost of delay instead of slipping back into the old prioritization game.