Regardless of political views, the WikiLeaks story surely did not go unnoticed. Like many of you, I am interested in the story as it affects us, as citizens of course, but it also has a potential implication on business. This motivated me to attend a dinner event featuring an amazing panel of experts:
- Daniel Ellsberg, Former State and Defense Dept. Official prosecuted for releasing the Pentagon Papers
- Clay Shirky, Independent Internet Professional; Adjunct Professor, Interactive Telecommunications Program, New York University
- Neville Roy Singham, Founder and Chairman, ThoughtWorks
- Peter Thiel, Technology entrepreneur, investor, and philanthropist
- Jonathan Zittrain, Professor of Law and Professor of Computer Science, Harvard University; Co-founder, Berkman Center for Internet & Society
- moderated by Paul Jay, CEO and Senior Editor, The Real News Network
The room was full of journalists, technology experts and others thirsty with knowledge. It was certainly one of the most diverse audience I have seen in a business event.
Why WikiLeaks matters? Why it doesn’t?
It was the focus of the event. It is not the focus of this post though. I have my own views of course on what is right and what is wrong; but I doubt that is what you are interested in.
Beyond the potential threat it represents on National Security or the impact of the transparency promised by the US Government and others, there are some aspects that relate more directly to Decision Management and some lessons learned or more precisely lessons to be learned.
I would like to explore the WikiLeaks phenomenon along 3 different dimensions. Here are the 3Ds:
- Strategic vs Operational Decisions
- Decisions under Uncertainty
- Effect of Transparency on Decisions
Strategic vs Operational Decisions
Some friends thought it was a stretch to consider WikiLeaks as a case story for Decision Management as we know it. They are right that we, in the Industry, spend most of our time focusing on tactical or operational decisions. By that I mean the decisions that do not matter individually but that occur in large volumes (think millions of transactions per day) although, aggregated, they might represent an important part of your business. Take for example the mortgage industry — in isolation, one borrower defecting on his/her mortgage is not tremendously impactful for a financial institution. Taken in aggregation, large volumes of defaulted mortgages can have a serious impact on the country’s economy as we are so painfully aware. Those decisions are quite different in nature from Strategic Decisions that typically apply to a very small number of instances but carry huge value. A typical example might be whether you should open a branch in a given location or what would be the right location for a new factory.
Steve Hendrick and Henry Morris, both industry analysts covering Application Development at IDC, have often talked about this dichotomy. Strategic and Operations Decisions are siblings in nature — all about Decisions — but they tend to present very different characteristics and to require different techniques or technologies. I encourage you to listen to one of their talks on the subject. They have put together a slide that put both on the map with those parameters (volume and value) in axis. This is a well-done representation.
The WikiLeaks story did rely vast amounts of information — on volumes of documents released to the public, actually more in their short existence than the Washington Post published in 30 years — but the key decisions that were made in the process were strategic in nature:
- Private Bradley Manning had to decide whether or not he should release the “Collateral Murder” video
- US Government had to decide to prosecute or not the leakers — as Daniel Ellsberg pointed out almost twice as many were indicted in this affair than by all previous Administrations in the past
- Corporations like Amazon and PayPal had to decide how to react to Joe Lieberman’s request to cease any business relationship with WikiLeaks
By design, Strategic Decisions are worth spending the time to think “humanly” — as in “having a real person think” of course: we are not judging any of the decisions that were made! –. Having an automated system for Strategic Decisions sounds both overkill and counter-productive. This is a space where BRMS would typically not apply. In that environment, Business Intelligence is the tool of choice to provide insight that can make the difference. The availability of Data is critical of course for the traditional approach. When you do not have data and no way to infer the data you don’t have, you will eventually have to trust your own judgment. But I believe that in most cases you can find some data that is relevant and informative if not perfect. This is the very reason why dashboards have become so precious in the past few decades. Granted those decisions are rarely easy, but the right information and skilled advisors certainly bring a lot of value to decision makers facing those impossible choices. The Art of Business Intelligence resides in the ability to smartly pick those data points.
In the past year or two, there has been an increased interest in those manual decisions. For example, the individual value of an underwriting decision made manually remains equivalent to other underwriting decisions. The Insurance and Financial industries have been investing over the last decade to increase the level of automation but there will always be some for which we just lack the formal knowledge. While we have humans making those decisions, we have been under serving this group compared to Strategic Decisions. They are not given the same level research. They are not given the same infrastructure. Recently, the concept of ad-hoc processes has gained some visibility. The tools are not there yet to provide as complete an environment as companies have for automated or strategic decisions of course but the ability to design a process on the fly as decision elements are uncovered is certainly a nice step forward.
Decisions under Uncertainty
What makes Decision Making difficult is obviously the fact that unknowns make the outcome hardly predictable. The panel talked about motivation and outcomes both for the Government and for the Corporations.
From a Government perspective, there is the clear mandate of National Security. No argument there. But officials also serve their constituency. Beliefs and affinities influence their judgment in matters that are certainly not clear-cut. Besides making the right and fair decisions today, the impact of those decisions might be of even bigger importance. As Daniel Ellsberg pointed out, the fate of Bradley Manning and others involved in this affair might alter forever the media. Assuming their conviction, reporters may not be allowed to keep their sources secret, at the risk of spending their whole life imprisoned for contempt. That adds a lot of pressure on present-day decisions.
From a Corporation perspective, multiple objectives are at stake when dealing with decisions relative to Wikileaks and transparency. Businesses are usually focused on the business outcome e.g. $$$. Mapping decisions to those bottom-line numbers is not always a simple game though. Take the Amazon decision to terminate their business relationship with WikiLeaks — supposedly as a result of Joe Lieberman’s call. From a business executive standpoint, beliefs may certain play a role but how easy could it be to anticipate the outcome of that yes/no decision? One one hand, caving to the Government request, like Peter Thiel put it, seems like the easier road. On the other hand, as Neville Roy Singham noted, foreign corporations may be reluctant to trust Amazon’s infrastructure for their business — hosting relationship turning into a hostage situation. The Canadian DNS company stood up to the demand, which might make them be seen as more reliable. I think we are too early to rule on what is the real financial impact. What I want to highlight though is the difficulty but also extreme value of being able to predict the effect of decisions on your bottom-line. If you can model your business in such a way that you can truly simulate those outcomes, possibly under a few likely scenarios, then you are in a better shape for making those make-or-break decisions. You might be surprised how much or how little has been done in this field depending on the industry. This was a take-away from the recent recession and I am glad that we continue to hear more emphasis on it.
Peter Thiel, as an entrepreneur, also emphasized the effect of pressure on decision makers. It is easy to say that you can and will stand up; when responsibility and accountability rests on your shoulders, you may act quite differently. He compared the Corporations caving to the Government to Government caving to the requests of terrorists. Provocative indeed, his claim does carry a little truth regarding pressure. A decision that you expect to be simple on paper may be quite different in person, when you face the reality of your actions. Not that I would ever consider a customer as a terrorist (per his analogy), but the real-life context allows you, and forces you, to take into account inputs in your decision-making. After crunching data, you may think that X is your threshold to approve applications — whether it is referring to income or risk score or anything else, there are and will always be exceptions that you will want to treat manually. What if the applicant is the son / daughter of one of your top customers? What if the person was the victim of an identity theft crime that altered unfairly the records and you just know it now? What if… This is yet again a scenario in which decisions informed by data and human judgment can be of greater accuracy and hopefully more profitable than purely automated decisions made in the vacuum. I like this idea of marrying both aspects in your decision-making process because you may not know ahead of time which one you will be able to trust. You mind might be overwhelmed by event and data can help get things organized as Carlos discussed recently in his post on catastrophic events. Data absent of human context may lack some flexibility when those unique exceptions occur… And those exceptions often make a difference in terms of turning an applicant into your most loyal customer or a vocal pest on public forums.
Effect of Transparency on Decisions
As you know I have been a big fan of Garter’s Pattern-Based Strategy. If you have not read it, I encourage you to look at the primer here. One of the 4 pillars is Transparency.
This whole WikiLeaks thing is obviously about transparency of operations of governments and other institutions. It is not about full transparency on everything, as we all agree, but it does raise the issue of whether it is okay to raise visibility on actions that are illegal or immoral and are secret in nature. This is a major issue: disclosure of such actions by the press resulted in impeachment procedures against President Nixon who resigned due to the near certainty of the outcome. Allowing whistle-blowers is a way to keep some safe-guards on the behavior of our leadership. This is not as easy question as this gray spectrum ranges from National Security to Democracy. We might all have different opinions on where the threshold should be. The aspect that I want to discuss here is the effect of transparency on decisions.
It is interesting for our purpose of Decision Management to realize that decisions might vary whether the outcome is to be public or secret. This has little to do with honesty. I am assuming that we are all trustworthy people! Visibility is a criteria that adds pressure to our decision-making process. Knowing that people are watching, safe-guards might kick in and alter your response to a stimulus. In some cases, it is problem — who has never felt surprisingly uneasy at an exam?
Peer-pressure and spotlights are not always bad though. Gamification theory states that those stimuli can actually have a positive effect on engagement and quality of decisions when channeled properly. I keep using the analogy of shooting darts for making decisions. I think it is quite appropriate indeed. In that case, assume that you are shooting darts constantly. With no crowd to show off and with no ability to measure your own performance, the game quickly becomes boring and, if you really have to continue doing it, it is likely that you will lose focus and throw roughly in the right direction with little precision. How can you outperform yourself when you do not have visibility on your performance?
Transparency, in my view, covers both aspects. Visibility into your own actions and their side effect as well as the ability for others to see your performance. Both are not always required of course. In some instances you want to share your results with your peers, because you seek their advice or because you want to show off. In all cases though, I think you need to measure your own performance.
This WikiLeaks panel was very insightful on the events. I enjoyed linking it to the discipline I am passionate about. Someone in the panel came up with a phrase we all found amusing “leak responsibly”. I will steal their witty remark for my conclusion:
Learn more about Decision Management and Sparkling Logic’s SMARTS™ Data-Powered Decision Manager