Operationalizing Decision Yield - Defining the questions
Several of our posts (like this overview) have talked about using "Decision Yield" to measure the efficiency and effectiveness of decision-making. But how do you about adding some "meat" to this concept? Well to use it you first need to develop specific questions that will allow you to measure your decision yield. Then you need to institutionalize the usage of decision yield in your business - are you just going to do a "decision performance audit" to see where you stand? Are you going to manage performance of a system by tracking decision yield over time? Are you going to make a decision yield for a core decision a standard KPI on your dashboard? Each of these will come up in the future but the core of this is the set of questions.
Now Decision Yield can only really be measured for a business decision area, not for a whole company or industry, so the questions must be built at a level that is specific enough to be compelling to a business decision maker. Ideally you want to create 40 to 50 questions that can be posed to mid- and senior-level managers from across a range of functions relevant to the area under investigation. Each question should take the form of a simple measurement of performance where the answer can be quantified or at least compared with some pre-defined standard. The answers to the questions will together generate an overall evaluation of a company’s performance across the dimensions of Decision Yield. Therefore, the dimensions of Decision Yield give helpful prompts as to the nature of the questions to be presented:
- Precision: Making decisions that result in the most profitable outcomes, for example, most reliably identifying and preventing fraudulent credit card activity, while minimizing the number of false positives
- Consistency: Interacting with customers the same way, regardless of the means of interaction, for example, ensuring that when customers are underwritten by an insurer, the same risk is priced at the same rate, irrespective of channel
- Agility: Being able to quickly adapt to changing business conditions, for example, putting an offer for your best customers immediately into effect in your call center when a competitor launches a new marketing program
- Cost: Providing the ability to increase the scale and scope of decision management with only an incremental increase in cost, for example, increasing the proportion of even complex decisions that are made automatically, rather than manually
- Speed: Returning decisions in as near real-time as adds value for both the organization and its customers, for example, reducing the time required to gather the necessary information and to underwrite a mortgage application
Beyond the simple categories, there are a number of ways in which a list of questions can be generated:
- Consider existing metrics used by companies in this industry. A good starting point, though unlikely in and of themselves to be sufficient
- Identify metrics used by equity and other analysts to rate the performance of companies. This is likely to be a rich source in "decision intensive” industries, such as banking and insurance, but may also come into play in others. It is also very valuable to highlight these particular factors when presenting the Decision Yield concept to senior management.
- Review annual reports of leading companies in the industry and pinpoint the metrics that they hold as being most important. These are likely to generate metrics which will resonate most from a competitive standpoint
For each question, the following additional elements need to be defined:
- To which managers should be these questions be addressed? The same question may be presented to multiple managers, and it doesn't matter if they give different responses, because the fact of different answers is of itself revealing
- To which dimensions of Decision Yield does a question apply? While a question might be prompted by consideration of one specific dimension of Decision Yield, it is possible that the response has implications across multiple dimensions
For each question you will want to define benchmarks against which you can evaluate performance. One approach is to define three levels that equate to three points on a simple scale of 1 to 5:
- The benchmark performance earning a score of 1 represents a "laggard”, or the level of performance expected from a very poor performer
- The performance earning a score of 3 represents a "follower”, or the level of performance expected from a company that is the middle of the pack
- The benchmark performance earning a score of 5 represents a "leader”, or the level of performance expected from a firm that is an innovator and at the forefront of decision performance
At the same time as defining the benchmarks, it is also necessary to define the overall weight that should be given to the response to this question. The importance of each question will fall into one of three categories:
- High: Performance in this category is critical to the firm’s competitive position and overall financial performance
- Medium: While important, performance in this dimension isn't necessarily "make-or-break”, as long as it is offset by good performance elsewhere
- Low: Strong performance is desirable, but not poor performance is not a "show stopper"
Tomorrow, an outline of a Decision Performance Audit and how you can use it to put a stake in the ground.
Postscript: Have you measured, or even thought about measuring, Decision Yield? If so, please contact me or James. We'd be delighted to learn about your experience in doing this.