09 October 2009

Back to Boeing

In June, I wrote my first hypothetical Prediction Market case study, featuring Boeing and its beleaguered 787.  Establishing a theme for subsequent blog posts, I posed the question:  What might be different if this company used internal markets?  The idea was to ask the same question about many organizations, exploring the myriad ways Prediction Markets can help solve today's business problems.

But today, this blog returns to Boeing for inspiration.

I don't mean to pick on Boeing, nor even fixate on them.  And I certainly don't mean to ignore all the other companies that have Prediction Market-worthy news headlines.  Indeed it's nearly impossible to flip through a single day's Wall Street Journal or New York Times without confronting an organization that needs improved decision intelligence.

But the latest development in the Boeing Dreamliner story provides a convenient segue to a topic I've wanted to address: the ROI of enterprise prediction markets.

From what I've seen in the marketplace (both first and second hand), Prediction Markets frequently face the same bottom line scrutiny as any other enterprise application, tool, or resource.  How much value will it generate and when?  Arguably, the business case should be extra tight when evaluating something new and unconventional.  And prediction markets fit the bill.  They are not yet widely adopted;  they stem from newfangled trends such as crowd-sourcing; and most egregiously, they challenge traditional management orthodoxies.

So it's no surprise that discerning corporate decision makers want to know what they'll get for their investment.  What is often surprising, however, is the way in which they approach the equation.

Here's what companies often say:
This seems very promising, but I'm not sure it's worth the cost.  What kind of returns can I expect?  What is the value of this new information?

Here's the cost-benefit equation those comments imply:
Value = benefit - cost
    or ...
ROI = benefit / cost
   where ...
        cost = price of Prediction Market solution + cost of internal time & resources
        benefit = value of the information (generated by the Prediction Market)
Not to get hung up on the math, but these simple equations are missing at least three variables:
  1. the cost of not doing
  2. the cost of alternatives
  3. the multiplying factor of the company's management "skill"
The cost of not doing
Here, I'm simply flipping around one of the original questions.  Instead of only asking what is the cost of doing something, sophisticated leaders also evaluate the cost of not doing something.  In other words, what is the risk of passing on a particular opportunity, or ignoring a particular problem?  Those risks should be considered, and considered as costs.  When those risks manifest into missteps, the costs become clear and public. (In the case of Boeing, a $1 billion write down.)  I've already written about this a bit in my last post, so I won't belabor the point here.

The cost of alternatives
Again, we return to the Boeing example.  Wednesday, the WSJ reported two interesting developments about Boeing's commercial airplanes division.  First, that production delays were not only plaguing the Dreamliner 787, but also the latest version of the 747.  Second, that the company has taken substantial measures to get a better handle on its production process.  This includes building a high-tech war room to more closely monitor the Dreamliner's progress:  
Vital to Boeing's plan for keeping the 787 on track as it starts building the 850 planes on order is a space center-style control room -- officially called the Production Integration Center.  One of the hub's wide glass walls overlooks the Dreamliner final assembly line, where the plane's body and wings come together. On the opposite wall, 24 big screens display information including overseas shipments of parts, urgent technical questions and even earthquakes around the globe, which could misalign factory equipment and cause delays.  Suppliers as far afield as Australia, Italy, Japan and Russia can call in through translators and show Boeing engineers in the center close-up images of their components using high-definition handheld video cameras. Robert Noble, Boeing's vice president of supplier management who runs the 24-hour center, says immediate, multimedia communications have eliminated the problem of often unclear email exchanges between distant engineers who work on opposite ends of the clock. "That takes days out of problem resolution," he says.
Additionally, the company has brought in-house previously outsourced parts of the manufacturing process, purchasing factories or taking large stakes in suppliers.  And finally, Boeing says it has added several project engineers to oversee key areas of the production line directly.

One doesn't need to be a rocket scientist (or a jet engineer) to recognize that these efforts are major financial investments. From acquisitions up the supply chain, to an expansion of middle management, to the technology, staff, systems, and secure pipelines that power the communications hub ... Boeing has accepted a very high price for improved decision intelligence.  Which, logically, means they have placed a very high value on course correcting the Dreamliner program.

Do they need to make all these investments?  As an outsider, I can't answer that question.  But one thing is clear.  The costs of alternatives to Prediction Markets (assuming that increased communication vehicles, added SME's, and tighter ownership are alternative ways of giving leaders better information) is high.  Very high.

Given that, my simple suggestion is that companies factor in those costs when evaluating Prediction Market solutions.  Once leaders decide that improved intelligence is a must have, not a nice to have - the actual costs of a Prediction Market are likely to pale in comparison to most alternatives.

One can't help wonder whether collective intelligence tools like Prediction Markets could have prevented the Dreamliner delays in the first place, saving the company billions in write offs, reputation losses, and expensive fixes.
The company's management "skill"

My final recommendation for the ROI equation focuses on the value of the market data.  Prediction Markets produce collective forecasts and other assessments, delivering them in the form of dynamic market prices.  Additionally, markets generate rich transactional data, which can help companies understand the who, where, and why behind the collective predictions.  When markets are successful, this data is both unique and valuable to corporate decision makers.  How valuable?  In my experience, that depends on two things.  First, the value of avoiding missteps and miscalculations like those discussed above.  This should be the first calculation.  And second, the actions of management in possession of the data.  This latter variable is simple yet often overlooked in the ROI calculation.  Unique information in the hands of passive managers may ultimately result in little value.  Intelligence in the hands of aggressive, knee-jerk type managers may result in some value or may destroy value.  Intelligence in the hands of the most skilled company managers can create significant value.  Just like any other piece of data that companies capture, value is neither guaranteed nor quantifiable until someone acts on it.
In conclusion, enterprise prediction markets should be held to ROI evaluations, as long as both costs and benefits are fully loaded.  In the case of Boeing, it seems a fair assumption that a relatively minor investment in collective intelligence could produce meaningful ROI.  Especially when weighed against the costs of not doing anything and the costs of alternative solutions.

07 October 2009

The Cost of Not Doing

In my quest to write hypothetical case studies on companies & organizations that need Prediction Markets, I've admittedly missed many easy opportunities. Just a quick mental scan of this summer's headlines produced several ripe candidates: The Yahoo!-Microsoft Deal, post-bankruptcy General Motors, the "closing" of Guantanamo Bay, the Cash for Clunkers program, etc. All are business or political endeavors whose fate is (or was) unknown, but whose future could have been predicted, potentially, by the proverbial crowd. If only the crowd had a Prediction Market, that is.

A Prediction Market is a crowd-sourcing tool that organizations can use for improved decision intelligence. Employees play the market ... executives get better information. These internal markets can be (and are being) used for forecasting, new product development, capital investments, and increasingly, project management.

But instead of jumping on these summer headlines and blogging up a storm, I got caught up with my own project management woes. Too many things to do, not enough time, yadda yadda.

In my case, there's no real cost to delaying my next blog post. But in the case of major initiatives in Corporate America or our government, there is plenty at stake. The costs of missed deadlines, inaccurate sales forecasts, budget shortfalls, marketing flops, rejected legislation, or failed mergers are significant, and scary. Which is why the relatively simple and inexpensive prediction market solution is so compelling.

As I've written before, Prediction Markets can't solve everything. But they can provide information that decision makers just can't get anywhere else. They can ask Yahoo! and MSFT employees for individual prognoses on the partnership, then aggregate it into actionable data. They can ask GM dealers across the country how many customers will return cars under the 60-day guarantee, ensuring the benefits exceed the costs. They can ask government officials from disparate branches the likelihood of Gitmo closing on schedule, uncovering hurdles and loopholes.

When the answers to these big questions can't be found among the usual experts or with the usual tools ... wise companies are beginning to trust the wisdom of the crowd. They're experimenting with prediction markets, wiki's, open innovation models, and the like. They're trusting the insights and ideas of their various stakeholders, not just their executives.

But many companies aren't. Many companies look at Prediction Markets and similar cutting-edge tools as unnecessary costs without guaranteed returns. This sounds interesting, but what's the ROI? That, the quintessential business question, can't be shunned. But it only takes a brief contemplation of the costs associated with big business blunders to entertain a new form of the question:

What is the cost of not doing?