The Means To The Ultimate End – Decision Analytics

Editor's Note: Today's post comes to us courtesy of guest contributor Chris Dobyns.

Have you ever had the experience of having said something pretty profound, but you didn’t realize it until someone pointed it out?  No, it doesn’t happen to me very often either, but it did last week.

Possibly as a result of some prior public prognosticating, I was contacted by the organizers of a future HR conference and asked to informally discusss what I thought might be the key trends for HR that were “just up the road.”  In an almost hour-long conversation, I reiterated some of the now familiar topics of big data, predictive analytics, applied behavioral science, data visualization and leveraging machine learning, artificial intelligence and other algorithm-driven techniques. At some point I apparently collapsed all those topics and unwittingly summarized them with the term “decision analytics”, which was declared by my listening audience to have surpassed the threshold of profundity (I think that’s an actual word).

 Decision Analytics – The Holy Grail of Business

Profound or not, I’ll acknowledge having something of an “epiphany” thanks to the listeners on my call.  Prior to that time, I’m not sure that I can say with a straight face that I fully realized the true purpose for all the various techniques and strategies that are undertaken to gather and analyze data. 

Turns out that all of that work is to allow someone to make an optimal (good) decision, by increasing the degree of certainty or confidence associated with the decision, while at the same time minimizing risk. Ostensibly, the “good” decision is then converted into an action which results in an outcome that approximates the fulfilment of a goal or objective. 

After making a mostly unnoticed “recovery” from my surprise epiphany, the conversation went on to further discuss that the beneficiaries of increased good decision-making in the future could be wide ranging – whether in the context of just business, or more specifically with regard to decisions affecting human capital management and pay and compensation. 

Good Decisions – A Many Splendored Thing

Armed with greater relevant data and more accurate and reliable predictions based on that data, the beneficial effects of improved decision-making in business are nearly limitless. With comprehensive information and an understanding of global economics, world events, technological progress, the geopolitical climate and a granular level of understanding the influence these factors have on the products or services that your business offers – what better decisions would be possible?  At my current employer, we openly and explicitly refer to this overarching goal as providing our leadership with “decision advantage.” 

With an improved understanding of the needs or wants of their potential customers and their likely behaviors, business can better appeal to or influence customer behaviors – and their buying decisions. Likewise, consumers will also continue to benefit from the ability to better understand and quantify their true needs and to make appropriate comparisons of available products and services, in order to make optimal buying decisions. 

Human Capital Management – The Decision Analytics Goldmine

The advent of improved decision analytics can also be anticipated to have substantial implications for the human capital management domain, to include issues affecting pay and compensation. 

With more relevant data to inform improved predictions, better decisions will likely be possible relative to:

• Optimizing the selection and fit of job candidates for existing vacancies
• Progressing toward the goal of near real-time labor market pricing
• Customizing pay/benefits delivery (type, amount, timing, frequency) optimized to individual employee needs, circumstances and career/life stage, to maximize productivity, satisfaction and engagement on the job
• Increasing the precision and effectiveness of pay and rewards budget planning and administration
• Improving the accuracy and effectiveness of workforce planning and workforce strategies

Beyond Human Decision-Making

Improved decision analytics would seemingly predispose good (or at least better) human decision-making.  However, decision analytics still cannot completely eliminate the influence of human biases, emotion, intuition and other irrational and cognitive limitations.  The likely solution hints at the future role of artificial intelligence in decision analytics and perhaps the wisdom of removing human beings from the decision-making process entirely.  Perhaps.

I recall the words of a former co-worker who liked to gently remind everyone to “always make good choices.”  The rise of nearly universal information availability in the next 3-5 years, combined with further advances in decision analytics, should put the goal of always making good choices . . . almost within reach.

Everyone probably has a different perspective.  What’s yours?

Chris Dobyns, CCP, CBP is currently employed as a Human Capital Strategic Consultant for the Office of Human Resource Strategy and Program Design for one of the largest U.S. intelligence agencies. The Office of Human Resource Strategy and Program Design is responsible for organizational effectiveness, personnel assessment, compensation and incentives, occupational structure, recognition and rewards, HR policy, human capital program design, implementation, evaluation and assessment and internal consulting. Chris has worked in the area of compensation for more than 35 years, and has been employed in various compensation-related positions by a number of large, private sector companies including, Sears, Roebuck, Arizona Public Service and Westinghouse Savannah River Company.

Original image "Good vs. Bad Decisions" courtesy of Chris Dobyns.

Comments

[E. James (Jim) Brennan]

Hear, hear, Chris!

Since every decision is by definition made from the Best Available Data at the time, we all make B.A.D. decisions all the time. Analysis of decision outcomes is a logical step to enhance the leverage of good BAD decisions. (Yes, that really does make sense.)

Identifying the types of decisions that have delivered optimum positive outcomes will indeed vastly improve the probability that the Best Available Data will be far more excellent in quality than ever before.

Same process could identify the contrary type, too. What are the characteristics of criteria that have generated highly negative outcomes? Then, avoid them.

Great stuff ...

[CIMT]

Thank you for your post. Keep it up.

[Rajiv Dalui]

Thank you for your nice post. Keep it up.

[Joseph M Thompson]

Chris,

Another great post.

Of course, implicit in the expression "relevant data" is the concept of accurate and current data - two attributes that cannot be taken for granted.

In addition, I never fail to be amazed by the propensity of people to have a wealth of relevant data and still make clear deleterious decisions - even knowingly deleterious decisions. (There's a book to this effect by David Maister called "Strategy and the Fat Smoker".)

Finally, there is a real likelihood of analysis paralysis that can accompany more data, even more analyzed data.

Overall, I suspect that there is some cosmic equation whereby the product of data x decision-making insights tends to be a constant!

[Chris Dobyns]

What? Even in the (future) face of hordes of relevant (and accurate and current) data you're worried that "other factors" could somehow encroach and snatch decision-making defeat from the jaws of presumably guaranteed victory?

Hmmm, sounds like you just may be advocating turning over most of the decision-making to the machines . . .

It worked out okay for Skynet, mostly.