INFORMATION TECHNOLOGY

Home > IT > Mysteries, puzzles, and business intelligence

---

Mysteries, puzzles, and business intelligence

June 5, 2007

What's the difference between these two business questions?

• "How many new customers did we attract last year?"
• "Are we attracting the right customers?"

The first, with apologies to best-selling author Malcolm Gladwell (Blink, The Tipping Point), is a puzzle; the second is a mystery. Gladwell explores the distinction in his New Yorker article, Open Secrets: Enron, intelligence, and the perils of too much information.

Gladwell's article should be essential reading for BI professionals. For as he illustrates, mysteries and puzzles involve distinctly different approaches to using information. And taking the wrong approach can often lead to disaster.

Puzzle vs. mystery: what's the difference?

Read our interview with Bob Woodward

A puzzle, in Gladwell's view, has a single, fixed answer. The more information you have to work with – letters in a crossword, for example, or numbers in a sudoku – the easier it is to figure out.

Watergate, he writes, is a classic example of a puzzle. Journalists Woodward and Bernstein broke the scandal by uncovering information that the Nixon administration tried to keep hidden.

A mystery is less definitive. In place of an answer, there are only interpretations. And the picture gets more confusing with each new piece of information.

"Mysteries require judgments and the assessment of uncertainty," writes Gladwell, "and the hard part is not that we have too little information but that we have too much."¹

Enron: too much info, and a classic mystery

Enron is an example here. As Gladwell illustrates, executives didn't hide their actions. Rather, they disclosed them all – including their creation of "special purpose entities" that were at the heart of the scandal:

"Everything [Wall Street Journal reporter Jonathan] Weil learned for his Enron exposé came from Enron, and when he wanted to confirm his numbers the company's executives got on a plane and sat down with him in a conference room in Dallas."²

The problem was that all this disclosure generated thousands of "numbingly complicated" pages of financial data – so many that no one could understand what was happening until it was too late.

In the end it took a combination of accounting experts, financial analysts, investigative reporters, the country's finest legal minds, and a great deal of hindsight to figure it all out.

Enron lessons for BI professionals

Gladwell's description isn't far removed from the way companies generate and use information.

No doubt your own company generates massive volumes of data that different groups use business intelligence to analyze, report on, and combine in different ways. But are they solving mysteries or puzzles?

With BI deployments growing in both size and complexity, their effectiveness (and your success) will depend in large part on your knowing the difference and configuring the system accordingly.

In Gladwell's thinking, solving puzzles means increasing the volume of information in play. But if you're dealing with mysteries, you'll want to develop your analytical capabilities and bring people together to collaborate on a solution.

Reports and queries can solve puzzles

Puzzles are based on simple questions about the state of the business, such as where products are selling or how well people are performing their jobs.

For example: a company might want to learn more about its pay-per-click ads. With BI, marketers can determine which days have the highest traffic and which ads have the highest conversion rates.

These are simple queries that business users can do themselves, or that you can easily automate. And a series of connected reports can bring users to increasingly detailed information. The answers can in turn lead to specific decisions or actions, the results of which can be easily tracked.

While analysis and scenario planning unravel the mysteries

Mysteries are more complex. In business, they revolve around strategic planning and priorities.

For example: if the question "How much are we investing in Asia this year?" is a puzzle, the question "Are we making the right investments in Asia this year?" has all the earmarks of a mystery.

This approach is most useful when considering future scenarios – a key responsibility of your senior executives. While it's impossible for them to know what will happen, it's critically important for them to think about what might happen, and determine how best to prepare for or respond to different scenarios.

Analysis, interpretation & collaboration are key

In these cases, there are a lot more what-ifs. And analysts, planners, and business experts must look at information from different perspectives and combine it in different ways. Interpretative skills are paramount.

Also important is a system that can support them with sophisticated analysis, planning and modeling capabilities, and scorecards.

Perhaps most vital, companies need a single source of trusted information. When people start analyzing and sharing knowledge, they know they're all working from and acting on the same numbers.

What to do when there's no "right" answer

Finally, since mysteries have no empirically correct answer, the only way for executives to know if they've made the right decision is to set targets – whether internally or externally focused – measure performance against them, and analyze the results in a continuous cycle.

Summary

Mystery or puzzle? Business intelligence can solve both with equal aplomb. But each line of inquiry requires a different methodology and information strategy.

Simple queries don't solve mysteries around company strategy or tactics. And complex analysis is overkill if users only want to understand regional sales or Web usage. Knowing which path to take can help you make the best use of your time, resources, and BI capabilities.

Find Out More Love a good puzzle? Read our interview with Bob Woodward Related reading: David Axson's take on managing for mysteries

---

Sources

¹ Malcolm Gladwell, Open Secrets: Enron, intelligence, and the perils of too much information. The New Yorker, January 8, 2007.

² ibid.