Understanding Business Processes with Coupling From The PastCoupling From The Past (CFTP) was one of the major advances inprobability theory during the 1990s. Now, we see increasinginterest in how to leverage its theoretical underpinnings tomaximize revenue. The product lifecycle has a particularlystraightforward application of statistical principles.The most fundamental concept in Coupling From The Past is theMarkov process. The important properties of a Markov process areAt any point in time, the process is in a particular current stateAfter a fixed interval of time, the process changes to its next stateThe probability of a particular next state depends only on thecurrent state, not on past statesHere is an example business process:There are five possible states:Vision,Design,Production,Maintenance, andObsolescence.The fixed interval of time is the quarter (i.e., January-March, April-June, July-September,and October-December). The arrows are labeled with the probability of a changeto a different state. The process might not change to a different state; in other words, thenext state can be the same as the current state. Those probabilities are not shown.For example, if we are in the Design state, with probability 0.6, thenext state is the Production state. With probability 0.1, the nextstate is the Vision state. Since probabilities sum to 1, with probability0.3 the next state is the Design state.Business RelevanceWe need to find the equilibrium probabilities of these states in orderto run our business. If the Vision state has high probability,our business stagnates as we contemplate product ideas with noforward progress. If the Design state has high probability,our time-to-market will be too long, and customers will gravitatetoward our competitors' offerings.We especially need to know the probability of the Production state,since it affects many aspects of resource planning. For example,if the probability of the Production state doubles, we may needto double our available factory floor space and double our numberof assembly personnel.The probability of the Maintenance state has perhaps the mostchallenging interpretation. Here, we are no longer seekingnew markets for the product, but we are still providing customersupport, repair, and upgrades. Clearly, this state should nothave a negligible probability, since this would drive downcustomer satisfaction. However, the probability should not be high,since our appeal to investors will decline if too few resources are dedicatedto growth.The Obsolescence state reflects discontinued products. The probabilityshould not be high since this would have a negative impact on branding, andproduce a perception of insufficient product focus. A reasonableprobability of Obsolescence is desirable, however, since it shows thatthe management team can make the hard choices to abandon productsonce they are past their prime.In Coupling From The Past, wewant to find the equilibrium probabilities of the states of a Markov