The Not-So-Intelligent Enterprise
Welcome to the not-so-intelligent enterprise, where too many of us live and work. Is there a way out?
In the late 1990s, two of Northern California's largest hospitals decided to merge. In public testimony at the state capitol and in scores of pronouncements and publications, executives from both hospitals vowed that the merger made economic sense. Two short years and hundreds of millions of dollars later, the two hospitals "de-merged," nearly forcing one of the hospitals into bankruptcy, disrupting services to thousands of patients, and forcing a state bail-out that caused a lot of red ink and red faces across the board.
More Software Insights
- The Critical Importance of High Performance Data Integration for Big Data Analytics
- Mobile DevOps: Achieving continuous delivery with multiple front ends and complex backends in Banking, Financial Services, and Insurance
- 2012 Excelian high performance computing maturity benchmark
- 9 Steps to Building a Business Case for B2B Integration
The sad truth is that the executives, to be polite, hadn't a clue what they were talking about when it came to the economics of the merger. They couldn't, because one and possibly both partners had financial systems that were completely undecipherable, misleading, and downright faulty. Whatever the top-line assumptions that the top dogs were operating under, the bottom line was that the bottom line was unknowable. That's just the financial side; there were other cultural, geographic, and common sense issues that spelled doom. In the end, the only certainty was failure, and the major question was when, not if.
Ignorance at Full Speed
Welcome to the not-so-intelligent enterprise, a place where unfortunately too many of us live and work. For those of us who believe in the power of the analytic enterprise, it's a sad fact that all the software and key performance indicators in the world can't analyze a bloody thing if the data, the basic assumptions, and the managers who use them are off track. In the case of the two hospitals, there were a lot of good reasons why even perfect software and analytics wouldn't have saved this ill-conceived conceit. But the fact that practice managers at one of the hospitals couldn't tell from one month to the next where the money was coming from and going to meant that any roll-up of those numbers by the bigwigs at the top would be just as faulty. Stupid is as stupid does, and stupid was doing a whole lot when it came to thinking the merger made any financial sense.
One of the reasons we know that the not-so-intelligent enterprise is both alive and, unfortunately, one of the dominant models for doing business today — perhaps more so in the U.S. than elsewhere — is that we know how bad underlying data can be. One great measure of how bad data can be is evident in the plethora of technologies, products, and initiatives that companies are deploying to clean up the mess. Behind all the three-letter-acronyms we take for granted today — ETL, EAI, EII (enterprise information integration), and so on — is the tacit admission that we have a tremendous need to improve how we measure what we do.
Customers are swarming around SAP's new master data management product — a technology that lets customers reconcile disparate corporate data under a set of standardized data masters — precisely because they know that they don't know how accurate their customer, partner, part, or employee data is from one system to the next. Informatica is now promoting universal data services — an architecture for unifying a company's disparate data integration and analysis functions into a leverageable, scalable, enterprisewide set of projects — precisely because the one-off nature of most such initiatives by definition means they are either overpriced or underperforming.
Managers glom on to Six Sigma, ISO quality programs, and every attempt at improving management and business practices precisely because they know they're lacking in the know-how to make things better before they get worse.
We know the enterprise isn't as intelligent as it should be because we know that virtually every ROI study lacks reliable "before" data that we need in order to truly measure the "after" success of a project. We know that executives don't use enough complex analytics because even the dumbed-down dashboards and dials that analytic applications provide aren't being deployed enough. And we can see that, fundamentally, not everyone seems to care: Executive compensation goes through the roof regardless of corporate performance, stocks trade at multiples that have no relation to their actual values, and the image of doing well seems more important than actually proving that things are as good as we say.
Cult of Personality
Why is the not-so-intelligent enterprise so prevalent, particularly here in the United States? I have a couple of pet theories. I think it starts with education. In Europe, even future poets have to study calculus in their teens, and future managers come equipped with an understanding of data and analysis that makes Americans look like a nation that walks around (with evident pride) carrying how-to books advertising that we really are the "dummies" our education system has made us into. I think it also has to do with the "great man" theory of American culture. We cherish cults of personality that elevate a few individuals — such as Jack Welch, the former head of GE, Bernard Ebbers of Worldcom, or Ken Lay of Enron — and ignore the contributions of the anonymous minions and real business processes that underlie dramatic success and failure: Who needs great analytics if you've got a great, intuitive CEO whom the herd loves to follow?
Other reasons? Human nature makes it very difficult to organize humans into groups of rational actors. We tend to fight to protect turf, not higher interests, and opt for the short-term gain because the long-term is so much easier to put off until tomorrow. We have little experience in collective action — this is the nation of the rugged individualist, after all — and we prefer proving our individual worth even if it means decreasing the collective worth and the value of others' efforts. We live and work in a society that wants — no, demands — that every hit be a home run and every home run be the tiebreaker, and sets little value in the incremental improvements that, over the long term, can make everyone a winner.
But finally, I think we as a society and as a business culture are moving at a speed few of us can maintain anymore. The real-time enterprise is real — plenty of businesses operate in a 24x7 environment — despite the fact that as humans we're biologically more suited to 9x5. We set expectations of performance for ourselves as individuals and corporations that we can't possibly fulfill, and then we fall back on a few great leaders who simply can't carry every ball in every play for every team. Most important, we ignore the tools — the analytics — that could help us manage these new models because we don't value their contributions or take the time to understand them. The result is a lot of bad decisions, based on bad data, which have a lot of bad results for companies, employees, investors, and customers.
Granted, many companies are trying to solve these problems, and I applaud all of you who read this magazine: By definition, you're part of the solution. But for the rest, I confess that for 20 years I've been watching businesses make mistake after avoidable mistake, and wondered why. Now that we have some impressive tools and products that have effectively removed the last good excuses for these errors, I think I know the answer. It's as simple as that all-too-prescient line from the immortal Walt Kelly and his equally immortal cartoon Pogo: "We have met the enemy and he is us."
Joshua Greenbaum is a principal at Enterprise Applications Consulting. He researches enterprise apps and e-business.