Minimize Risk with Data Analytics

Are you climbing the ladder? What is your company’s personality? Don’t forget time is money! Using metaphors is second nature to us, and there is certainly no shortage of them in business. While these metaphors quickly establish a common ground of understanding between different parties, they can actually interfere with decision making. One more reason that decisions must be quantifiably justified.

In 2012, researchers from the University of Toronto and the University of Michigan played with this idea. Considering the metaphor that “something smells fishy” they exposed study participants to piscine smells. They then had the subjects participate in trust based games. Surprisingly, there was a drop in trust and cooperation.  It even worked in the other direction, where participants who were primed to be suspicious were more easily able to identify fishy smells and could even detect low concentrations of fishy smells better than others.

These sorts of subconscious biases can be found all throughout our decision making processes. Every psychological study that speaks to this is just one more piece of evidence of how necessary it is to be able to describe why a decision is being made. Arbitrary gut feelings (sometimes) work in instances where there is not enough time to properly work out a course of action. However, when not pressed for time, there is no reason not to make every major decision with as much data and analysis as possible.

Data visualization, of which we spoke last week, is a major part of this. It allows you to quickly and efficiently interpret the data you collect. But visualization is not the entirety of analytical decision making, and it cannot answer every question.

 

Risk of a Bad Decision

The situation is further complicated by the realization that there is more than one type of “bad” decision. When most people speak about business, they think of only one type of negative decision; pursuing a course of action that should have been avoided. This is the typical “waste of money” whether it is a poor investment, a bad hire, passing an item that should have failed, or a process inefficiency.

However, we all too often forget about missed opportunities. When you decide not to make a leap you should have, that should also be considered a waste of money. People naturally feel more pain at a loss than a foregone gain (hello prospect theory!) but both must be avoided if our companies are to grow and succeed.

Both of these decisions are therefore bad; which is worse can only be determined by an organizations risk management process. In order to minimize risk and outsize influence from certain factors, organizations must have an analytical process in place and insist that it be followed for all significant decisions. While the human element has its place, recognizing its weaknesses and finding things that supplement them is a must.

 

Which to Choose???

Sound risk management and good decision making requires data and a decision making process based on the analysis and interpretation of this data using a number of methods; probably the most commonly known among them in business circles is A/B testing.  A/B testing is great for businesses small and large. It allows for rapid experimentation and data collection with minimal investment. Everything from a minor ad buy on LinkedIn or Facebook on up to an international branding campaign can be A/B tested. The difference is only in the scale.

For the uninitiated, A/B testing involves finding specific variables to test in a comparative study to find the best option. For instance, if you were buying ads on LinkedIn with a header text and body text, you could A/B test the header text by leaving everything else alone and making variations to that. And A/B testing does not have to be constrained to two variables. It can be used for numerous factors, so long as any potential external factors are accounted for and the results are closely tracked. This is where the value of a trained analyst come in.  Many a small business has gotten lost in the results of their own A/B testing, unable to differentiate which factor is causing which result.

 

Cost of Poor Quality

Another key application for decision making in business is in the area of quality assurance / quality control.  When it comes to making decisions regarding quality, the two types of risks associated with a bad decision can become even more critical.  In one sense, costs of assuring quality to ensure customer safety and satisfaction can lead to being a bit too conservative; rejecting a perfectly good product or spending too much time and money beyond what is necessary.  This is often referred to as “producer’s risk”, since the producer of the product is incurring additional, unnecessary costs due to poor insights.

Conversely, is the risk associated with passing on a bad product.  In some industries, this could have serious consequences, possibly causing harm to the customer.  As such, this sort of risk is commonly called “consumer’s risk”, as the consumer is the one assuming an unintended risk associated with a bad product that escaped quality assurance controls.  Here too is due to poor insights, where ultimately the risks could be minimized with good data analytics.

 

Risk of Poor Data Analytics

Risk of bad decisions may prove to be a serious issue for smaller organizations without dedicated analytics personnel. Compound this with the fact that many times the data from this testing comes natively through third parties. For instance, when running social media ads, the data is of course reported through the platform itself. This is not a major impediment – and the data can often be exported – but it is one more complication for those already struggling to implement effective analytics solutions. Also consider the validity of the data.  Without a solid data strategy, even good analytical techniques can lead one astray when sources with bad data.  As the old saying goes, “garbage in – garbage out”.

Another old saying is “more is better”. While increasing the volume of data while maintaining veracity increases the confidence with which you can make decisions, the costs associated with collecting and analyzing more data than necessary is a bad business decision in and of itself!  Also keep in mind that not every data point will be relevant! This is the heart of data analytics; making decisions with confidence using the right amount of the right data.

Next week we will be staring into the crystal ball while we discuss the use of data analytics for business forecasting. This is critical for your business to recognize potential opportunities and threats as they appear on the horizon, rather than when they are right in front of you.

Sharing Is Caring!