Agile Risk Management – Complexity and uncertainty are the new norm

When com­pa­nies focus on min­i­miz­ing risks, they simul­ta­ne­ous­ly reduce oppor­tu­ni­ties. To seize oppor­tu­ni­ties, we need cre­ativ­i­ty and courage. Courage requires con­fi­dence. Even though Ed Cat­mul­l’s quote is from a 2008 pod­cast, it is still rel­e­vant today. Risk man­age­ment must change so entre­pre­neurs and employ­ees can leave their com­fort zones and try new things. This is the only way to exploit oppor­tu­ni­ties effectively.

Everyone is required

Every­one is required to obtain the best pos­si­ble infor­ma­tion. Every­one pro­vides the exper­tise for their work area to iden­ti­fy and eval­u­ate risks and oppor­tu­ni­ties and make bet­ter deci­sions. Com­mu­ni­ca­tion is a crit­i­cal fac­tor in this process. Before we can effec­tive­ly share our knowl­edge with oth­ers, we need a com­mon under­stand­ing and lan­guage that makes mis­un­der­stand­ings avoidable.

Prob­a­bil­i­ty terms such as “low, medi­um, or high” are stan­dard in risk assess­ments. When describ­ing impact, terms range from “neg­li­gi­ble to cat­a­stroph­ic.” But what does it mean? Every­one under­stands some­thing dif­fer­ent by it. Thus, mis­un­der­stand­ings are bound to occur, lead­ing to wrong decisions.

The fig­ure shows a typ­i­cal dis­tri­b­u­tion of respons­es from work­shop par­tic­i­pants to the ques­tion of how they define the prob­a­bil­i­ty in per­cent for the terms men­tioned on the left.

The results sug­gest that more than the implic­it dif­fer­ences in per­son­al def­i­n­i­tions are nec­es­sary to com­pare assess­ments. To avoid this, use num­bers and com­pa­ra­ble scales.

Probabilities and opinions

Prob­a­bil­i­ties are expressed in per­cent­ages from 0% (impos­si­ble) to 100% (cer­tain). Most peo­ple have learned that the per­cent­age val­ue indi­cates the rel­a­tive fre­quen­cy of occur­rence of an event. For exam­ple, the prob­a­bil­i­ty of rolling a 6 is one in six or 16.6%. Of course, we don’t know this for arbi­trary events because we need more infor­ma­tion. Here is where Bays’ con­cept of prob­a­bil­i­ty comes in. The per­cent­age val­ue is a mea­sure of our belief. We express our con­fi­dence in a num­ber com­pa­ra­ble to oth­er assessments.

While deal­ing with prob­a­bil­i­ties is often a chal­lenge for us, there is one val­ue that we are very famil­iar with the mon­e­tary value!

What does it cost?

What is the cost or prof­it if a par­tic­u­lar event occurs? Since it is not easy to give a con­crete sum, we define an inter­val in which the actu­al val­ue is like­ly to lie. Using mon­ey as a scale makes esti­mates of impact com­pa­ra­ble. Com­mu­ni­ca­tion of beliefs and esti­mates is more explic­it in this way.

Is there an advan­tage to express­ing beliefs and esti­mates in num­bers and com­par­ing them? The answer is yes because, in this way, expe­ri­ence and intu­ition become vis­i­ble and valu­able for decision-making.

The value of intuition

Our brain can rec­og­nize pat­terns in com­plex con­texts. This abil­i­ty is also known as intu­ition or gut feel­ing. It enables us to make quick and unerr­ing deci­sions where ratio­nal thought and analy­sis take too long. How­ev­er, there are also lim­its. In terms of evo­lu­tion­ary his­to­ry, humans have lived most of the time in a sta­ble sit­u­a­tion over a long peri­od. Our ances­tors had enough time to learn to use their intu­ition in such a way that it deliv­ered results in pre­cise­ly that situation.

If exter­nal cir­cum­stances change, our gut instinct deliv­ers results cor­rect­ly in the pre­vi­ous sit­u­a­tion but may be cat­a­stroph­i­cal­ly wrong in the changed situation. 

We trust our gut instincts so that we don’t hes­i­tate to apply poten­tial­ly vital deci­sions. Ques­tion­ing or ana­lyz­ing could mean the dif­fer­ence between life and death for our ances­tors in the wilder­ness. Today, ath­letes use this skill in team sports like soc­cer or bas­ket­ball, where quick reac­tions are required. Prac­tic­ing moves and ana­lyz­ing past games helps to make the right deci­sion intu­itive­ly and with­out long thinking.

Using intu­ition and expe­ri­ence in deci­sion-mak­ing in projects or the com­pa­ny is famil­iar. How­ev­er, the sit­u­a­tion has changed. The rea­sons are expo­nen­tial­ly advanc­ing inter­con­nect­ed­ness, dig­i­ti­za­tion, and an increas­ing­ly unsta­ble polit­i­cal sit­u­a­tion. As a result, mar­kets are chang­ing much faster. Com­pared to sports, it is as if new rules apply to every game, but the play­ers are not informed about them.

This new sit­u­a­tion is the same for all mar­ket par­tic­i­pants, so com­pa­nies have many new oppor­tu­ni­ties to posi­tion them­selves and effec­tive­ly use the con­stant change.

Binary thinking

The ques­tion is: What is the ben­e­fit of intu­ition for projects and busi­ness deci­sions in a com­plex, fast-chang­ing world? — For­tu­nate­ly, the time for deci­sions is longer than in sports or the wilder­ness. Fast means faster than oth­ers. Since many com­pa­nies still have very long deci­sion-mak­ing process­es, even a tiny improve­ment can pro­vide a sig­nif­i­cant advantage.

If you don’t look at intu­ition in iso­la­tion but com­pare the results with oth­er experts, it’s like dif­fer­ent pieces of the puz­zle com­ing togeth­er to form a more com­plete picture.

Of course, it is not help­ful to insist on your point of view. Chang­ing one’s mind is eas­i­er when bina­ry think­ing, i.e., right and wrong or yes and no, is replaced by any num­ber of gra­da­tions in between. In this way, one’s assess­ment is not imme­di­ate­ly incor­rect, but per­haps only a few nuances less correct.

Risk aversion

But how is it that dif­fer­ent experts eval­u­ate the same facts dif­fer­ent­ly? One pos­si­bil­i­ty is men­tal bias.

For exam­ple, every­one has an indi­vid­ual atti­tude toward risk. Some peo­ple find it easy to take risks, while oth­ers shy away from them. The result is cor­re­spond­ing­ly dif­fer­ent assess­ments. These per­son­al men­tal bias­es are bal­anced by com­par­ing and com­bin­ing them with the results of others.

Oth­er pos­si­ble rea­sons are dif­fer­ent expe­ri­ences, lev­els of infor­ma­tion, and perspectives.

Information and consistency

The rat­ings show con­sid­er­able dif­fer­ences depend­ing on how informed a per­son is and how the respec­tive infor­ma­tion is weight­ed. For exam­ple, a piece of infor­ma­tion is rat­ed too high­ly because some­one has read a suit­able news item on the Inter­net on the same day. This phe­nom­e­non is called avail­abil­i­ty bias. 

Experts whose assess­ments and eval­u­a­tions play a role in deci­sion-mak­ing can learn how to improve their per­for­mance by ana­lyz­ing and, if nec­es­sary, adjust­ing their thought processes. 

Con­sis­ten­cy with avail­able infor­ma­tion also plays an impor­tant role. For exam­ple, we quick­ly get dis­tract­ed by case-spe­cif­ic infor­ma­tion and ignore avail­able, rel­e­vant sta­tis­ti­cal infor­ma­tion, such as com­mon­ly known prob­a­bil­i­ties for events or sam­ple sizes. This ten­den­cy is called the base rate fallacy.

The bet­ter the avail­able infor­ma­tion is used, and the more experts are will­ing to chal­lenge and cor­rect their esti­mates based on new infor­ma­tion, the bet­ter the results. These results form the basis for deci­sions; a sol­id foun­da­tion makes good decisions.

But how do we know how good our esti­mates are and thus the basis for decisions?

Feedback

In agile prod­uct devel­op­ment, we often ask users for feed­back. Based on the feed­back, we review deci­sions and cor­rect them if nec­es­sary. In this way, we learn more about the com­plex envi­ron­ment in which we want to present a suc­cess­ful solution.

When esti­mat­ing the prob­a­bil­i­ties of ran­dom events, this pos­si­bil­i­ty does not exist. Nev­er­the­less, it is nec­es­sary to receive some feed­back so that con­tin­u­ous improve­ment of results through learn­ing is possible.

The Brier Score, pro­posed by Glenn W. Brier in 1950, can help. The basis for cal­cu­lat­ing the Brier Score is the pre­dict­ed prob­a­bil­i­ties for the occur­rence of events in a cer­tain peri­od, e.g., in the next year. At the end of the peri­od, we know whether an out­come has occurred (1) or not (0). The Brier Score is updat­ed to pro­vide a val­ue between 0 and 1. The best val­ue is 0, and the worst is 1. 

Using the Brier score, the qual­i­ty of the esti­mat­ed prob­a­bil­i­ties can be objec­tive­ly checked and com­pared with pre­vi­ous results.

The esti­mate of the impact of events that have occurred is com­pared with actu­al costs or rev­enues. In future assess­ments, we can apply this knowl­edge to achieve bet­ter results.

What a coincidence

Com­plex­i­ty also means it is insuf­fi­cient to inde­pen­dent­ly con­sid­er prob­a­bil­i­ties for events and effects. It isn’t easy to intu­it com­bi­na­tions of ran­dom events. Sta­tis­ti­cal meth­ods can help but require spe­cial knowledge.

An alter­na­tive is to use quan­ti­ta­tive mod­els, such as those cre­at­ed with Excel or Jupyter Note­book. A sim­u­la­tion with ran­dom­ly gen­er­at­ed val­ues for the occur­rence and impact of events deter­mines the prob­a­bil­i­ty of loss­es or gains based on the com­bi­na­tion of all events and ran­dom val­ues. This method is called Monte Car­lo sim­u­la­tion. The results can be pre­sent­ed as a loss-exceedance curve and form the basis for fur­ther decisions. 

Such mod­els help to under­stand the phe­nom­e­non of chance better.

Limitations of models

Mod­els enable the ana­lyt­i­cal con­sid­er­a­tion of many inter­twined risks and oppor­tu­ni­ties. How­ev­er, each mod­el has its lim­i­ta­tions. For exam­ple, some risks are unac­cept­able because they con­flict with eth­i­cal and moral prin­ci­ples. In these cas­es, it is help­ful to use eth­i­cal and moral behav­ioral guide­lines as a basis for informed decision-making.

Conclusion

Risks and oppor­tu­ni­ties are two sides of the same coin. A rapid­ly chang­ing world and increas­ing­ly com­plex mar­kets offer new oppor­tu­ni­ties for com­pa­nies. It is nec­es­sary to take risks to take advan­tage of these oppor­tu­ni­ties. Com­pa­nies that adapt quick­ly to new sit­u­a­tions have a deci­sive advan­tage. Good deci­sions are essen­tial for this. 

Agile risk man­age­ment pro­vides a frame­work for a mod­ern, inclu­sive approach to prob­a­bil­i­ties, infor­ma­tion, uncon­scious knowl­edge, and expe­ri­ence to improve assess­ments, eval­u­a­tions, and deci­sions continuously. 

Agile risk man­age­ment puts peo­ple at the cen­ter of entre­pre­neur­ial action. The process is indi­vid­u­al­ly designed and con­stant­ly adapt­ed to new circumstances.

Those who can­not find an answer to the ques­tion of the ben­e­fit of the cur­rent risk man­age­ment sys­tem in the com­pa­ny should take action now. Those who still think risk man­age­ment is unnec­es­sary can only be wished good luck.