This article originally appeared on the BeyeNETWORK.
Kierkegaard once observed that people live life forward, but they understand it backward. IT and business managers have learned a lot about what has and hasn’t worked with their business intelligence, CRM, data quality and data warehouse projects. They’ve taken steps to avoid the errors of those who have gone before them, researching best practices, retaining experienced experts, engaging business stakeholders during planning and establishing success measures.
Company cultures and human nature being what they are, the more we educate ourselves, the more we learn what we don’t know. And so it goes with business intelligence (BI): faulty assumptions remain and old habits die hard. Those of us who have been in the data warehousing world for the past 20 years have seen our share of slipups in what I call “The Old Standbys” of business intelligence success. For instance, business analysts still forget the difference between business requirements and data requirements. Metadata remains an afterthought. Maintenance and support efforts prematurely fizzle. Data quality is so much lip service. And we all know managers who insist on treating – and funding – the data warehouse as a technology platform and not as a business solution.
These issues might sound familiar to you, and you might be in the process of working on them. But as the business intelligence industry evolves, new challenges evolve along with it. Thus, our goal should be to keep one eye on today’s challenges, and one eye (or at least our peripheral vision) on tomorrow’s potential “gotchas.”
In the past year, we’ve surmised a new crop of gotchas – barriers to BI progress – and we’ve noticed that no company is immune. Some of these firms have succeeded in securing management support for business intelligence, some haven’t. The gotchas subsist in both new projects and incumbent BI environments, in mid-market firms and in the Fortune 50. Most of these companies understand the basic rules of data warehouse success: be requirements-driven, keep it warm with the business, develop and deploy incrementally, cultivate fresh skills. They are launching their BI programs with more rigor than those who preceded them.
Nevertheless, the following seven phenomena threaten to sabotage startup and mature BI environments alike. What’s interesting is that we’ve encountered them at companies who consider their BI environments mature. Hopefully this list will serve as a checklist of warning signs for your 2007 planning and implementation activities:
- Businesspeople still pick up the phone. Even companies with robust enterprise data warehouses and tools that span the BI life cycle are falling victim to this one: when people need new information, they call someone they know and ask for it. It’s a business process glitch as well as a culture issue, so much so that we’ve built it into our scorecard methodologies as a metric. The extent to which business users call someone or send an e-mail request (as opposed to running a report or accessing online metadata) is the extent to which the BI infrastructure still needs refinement.
- Executives are in data quality denial. Everyone always thinks their data is better than it really is. I’ve never been brave enough to tell a client to “prove it,” but I have watched data profiling proofs of concept where the results leave the VP of Marketing slack-jawed and have the CIO scheduling emergency working sessions. More than once, the response begins with, “No wonder...” followed by an explanation of the failure of a key business program or IT project. To wit, a recent exclamation by BI business sponsor in the finance department: “No wonder that monthly KPI report said we had almost double the number of business customers we actually do!” Uh oh.
- IT is doing the wrong kind of spending. Trust me on this one: your IT management is still more focused on platforms (their cost, their maintenance, their uptime, their outsourcing) than it is on corporate data. And the quality of IT spending these days leaves a lot to be desired, with many investments focusing more on short-term fixes rather than long-term solutions. The best CIOs are those who see the rest of the company as the customer. These CIOs constantly measure and proselytize the business value of their IT initiatives.
- Solutions fail the “how valuable is it?” test. Many IT departments still run “loose” budgeting cycles or are excused from developing corporate-sanctioned business cases. While this might initially seem like a luxury for budget-rich CIOs, it brings bad karma down the line. All it takes is a C-level business executive on a cost-cutting spree to ask the dreaded question, “How much have we spent on that data warehouse, and what have we gotten from it?” When you have to look a client in the eye and explain that “the total investment in your data warehouse has exceeded any potential business value it will deliver,” it’s usually a bad day.
- Competency center confusion. The goal of a competency center is to accelerate the delivery of a core capability by leveraging skills, processes and reusable assets, ultimately decreasing the cost of development and maintenance. Competency centers should be institutionalized as a service to the business, and there should be consensus around their value. However, executives are failing to differentiate information competency centers from integration competency centers from BI competency centers from data quality competency centers. End users are baffled by the organizational ownership boundaries of the various competency centers. The result is often…er…how to say this diplomatically…oh what the heck: Incompetence. (On the horizon: I expect “governance confusion” to be a bullet point in next year’s “Gotchas” article.)
- Business processes that aren’t aligned. “Business processes create and consume data. They have owners,” explained Kevin McDearis, CIO of CheckFree’s software division. McDearis was talking to us for our book about customer data integration. “It just makes sense that business processes should drive everything.” But many companies still don’t know which systems generate the data and whether that data is created in more than one place. So Finance doesn’t know whether people’s withholdings match what’s in the accounting system, Marketing doesn’t know who’s responded to a campaign, and Sales can’t accurately assign territories. Integrating master data with key business processes is still on most companies’ wish lists, if not in their 2007 planning. And the larger the company, the more difficult this problem is to solve.
- Analyzing customers but still not really knowing them. As at least nine different CRM project managers I know can assure you, just because you can get a glimpse into a customer’s behaviors via business analytics doesn’t mean you’ve achieved customer intimacy. For instance, a state government we work with leaves 400 bytes of space for case workers to type notes about welfare and back to work constituents. This unstructured data is rich with promising content – a food stamp recipient turned out to be a deadbeat dad, and tracking him down could save the state a ton of money – but the agency’s BI tools aren’t equipped to parse the data and to find the relationships. This isn’t only a tool issue, it’s a data issue too. Getting unstructured data onto the radar of data warehouse and BI teams is timely and important. I know three separate CRM teams who have put “support unstructured data” on their 2007 to-do lists.
Since growth, customer retention, and shareholder value will always be “burning platforms” for corporate executives, it’s time for us to begin relating data to these and other strategies. Clean, managed, and integrated data should thus become an enterprise goal. This has interesting implications for the careers of BI and data warehouse professionals. Indeed, managing data as a corporate asset might become the missing link in IT evolving from “cost center” to strategic business partner.
Jill Dyché is a partner co-founder of Baseline Consulting, a technology and management consulting firm specializing in data integration and business analytics. Jill is the author of three acclaimed business books, the latest of which is Customer Data Integration: Reaching a Single Version of the Truth, co-authored with Evan Levy. Her blog, Inside the Biz, focuses on the business value of IT.