Via LinkedIn : I was watching a TED Talk by Teman Cooke called “The scientific method is crap”. Teman Cooke, Ph.D. is a theoretical physicist and explains a “Cycle of Scientific Thinking” (shown to the left) as an interesting alternative to “the scientific method”. This presentation resonated with me in particular because what he explained is how I approach problem solving and work with fellow employees and clients in trying to better understand why business performance is not what it should be.
The process usually begins with a question about can you help me with “insert topic here”? Topics range from strategy not working, too much time in the business, high employee turnover, improve profitability, stagnating sales to ‘I am not sure what to do next’. My immediate follow up is to ask why they think this matters so much.
At this point I am trying to establish what they are ‘observing’ that concerns them. As they tell me what they believe is wrong, I clarify it a bit further to ask, what they think the situation should be? By studying what they ‘observe’ (the ‘as is’) and comparing it to what they believe it ‘should be’, we begin to define the problem. Often variables upon themselves are not the problem, it’s the relationship between variables that is the problem. By means of example a key variable might be sales and I have never heard of problems with increasing sales by itself. Increasing sales by itself has never been nor will it ever be a problem. What follows can be the insight to the problem we are seeking; 1) “and I am losing money at a fast rate”, or 2) “and I don’t seem to have enough cash on hand”, or 3) “my revenue line does not seem to follow my sales”. So with increased sales, we expect some else to follow suit. It is when this ‘follow suit’ does not coincide to our own thinking, we surmise we have a problem.
So now that we have an observation, stated what the expectations should be and determined why it is a problem, I begin to ask more questions and understand what the process is;
- How sales are booked,
- how pricing is established,
- how costs are incurred,
- how and when collections are done,
- when and how expenses are made,
- plus a whole slew of other topical items
With these questions I try to determine the end to end process of generating leads, converting them to sales, then sales orders, then order processing, product service development, product service delivery and finally on to post acquisition support and service. The questions asked remain within the scope of the topic being studied. Often, if in doubt about some information, I will capture it because I can always determine its relevance later on when the model is developed.
With all the process information, timing, captured data points and financial reports available, I will develop a model that maps timeframes, the process, products and services produced to sales, revenues and costs. The model will incorporate some levels of regression analysis, timing and mix. The model is often quite complex involving multiple sheets thereby creating a multi-dimensional view of the operations. One of the main outputs of the model is to be able to forecast expected outcomes given a change in input variables. It also should be able to demonstrate how much the output will change for a unit change in inputs. I call this finding the points of leverage. So back to our example we may find that in terms of impact to the bottom line, some small changes in one product could have more of an impact on net income than a large cost reduction in another product due to the relative mix of sales experienced during the study period. Likewise reducing processing times may have a greater impact on cash flows in the short term than any cost reduction could. After all – we all know time is money.
The next step after model development is to test the model to see how it can explain the results we have been observing. I have found that most modeling has been very accurate and can definitely be used to conclude we have tendencies occurring that can be monitored and tracked. Sometimes I need to investigate further to find out where additional variability may exist. Most times I do find plausible explanations which are further supported with documentation and often labelled “special deals/items” which are the ‘out of policy’ decision being made in consideration of a particular situation. We are able to produce cause and effect statements which are backed up by observation, records or other ‘paper-work’. This step usually provides surprises to people in that they did not expect the causal relationships that begin to emerge in this type of study.
After we agree on the observations, the cause and effect relationships identified, we can then start developing some predicative outcomes given changes in input variables. We usually change the lowest implementation cost and controllable items first. This gives us the “biggest bang for our dollar” scenarios. We then follow this with lower implementation cost, less controllable (ie: 3rd party intervention required) options, and so on. The issue with control is that some items are very controllable by the company, others are not. Engineered costs for instance tend to be very controllable, whereas contractually established cost may not be so controllable. Market price and customer buying volumes are less controllable as well. Price setting is very controllable, buying volumes less controllable. We always pick most controllable items first, since they are most doable first and usually do not require permission to undertake.
Testing / Checking
Because the model identified the variables that impact results most, as a by-product, we also developed a basis of measurement, or leading indicators, that we can monitor ongoing. The beauty of leading indicators is that if we properly control our inputs, then the outcomes become more predictable, repeatable and reliable. Monitoring the inputs and outputs gives us a systemic way of measuring our processes from start to finish with reliable outcomes that should not surprise us. It is at this point we start the cycle of ‘observation’ over again and will update our model as new learnings begin to surface.
Applications of Cycles of Scientific Thinking
I have successfully used this process in pricing, product direction, business performance improvements, product margin analysis, profitability analysis, process trouble shooting, value proposition development, manufacturing process improvements and just about anything that you observe and can assign a value to. It’s a great tool to take the emotion and bias out of decision making. It is not to say that this tool is the only basis for decision making, you should always keep your company’s’ goals and objectives in the forefront of the thinking process to ensure the results help you achieve your success and stop you from pursuing low cost implementations that will hamper your success. You should also be prepared to consider that one outcome may be that you need to reassess your key strategies because the data may suggests that current strategies are not achievable given your current environment. This is often found in trying to bolster a product line that a new technology is beginning to replace, or keeping a service going because that is what we always done.
About the Author
Rudy Fischer is CPA,CMA and a partner in RK Fischer & Associates which is a business consulting and coaching firm who helps owner operated businesses in Canada maintain or regain profitability.
Practicing as a Chartered Professional Accountant , Rudy offers clients help in matters of Financial Wellness, Budgeting & Cash Flow Analysis, Financial Accounting, Financial Management and CFO / Controller services that delivers both intelligence and results to their business either on a contract or interim basis.