“In Store for ‘09: Prognostications for the New Year” by Jill Dyché was originally published in the January 15, 2009, issue of the TDWI e-newsletter TDWI FlashPoint. This article is being reprinted with permission from TDWI. For more information about TDWI, please visit www.tdwi.org.
Jill Dyché Baseline Consulting
If you ask me, there are two kinds of people in this world: those who read annual predictions lists and those who admit they read annual predictions lists. Either way, we’ve been seeing several evolving trends in the business intelligence and data integration worlds, as well as some phenomena that are about to bust wide open. Herewith, my predictions for 2009:
1. Software as a service (SaaS) will continue to challenge rapidly commoditizing ETL, data cleansing, and data enrichment functions. Companies are giving these so-called “infrastructure” technologies a second look to see how they can drive efficiencies. I’ve always maintained that the value of IT isn’t in the systems it builds but in the information it delivers to the business. SaaS enables companies to quickly deploy functionality, often in a fraction of the time it would have taken them to procure specialized software, hire and train internal staff, gear up installation planning, and deploy the necessary functionality. IT executives, especially those in the mid-market, are quickly realizing that they don’t need an army of people or a large budget for specialized toolsets if their functional requirements are well defined.
Companies offering SaaS capabilities for the first time, such as Acxiom, understand their customers aren’t buying cloud computing; they’re buying specialized hosted applications.
2. Data governance will abandon organization and embrace process. As business managers take steps to formalize the policy making and oversight of enterprise information, their knee-jerk first step is to call a meeting. People from many organizations, including IT, convene and nod their heads when the inevitable “data as an asset” notion is raised. However, when the conversation turns to data definition, ownership, and correction methods, consensus turns to strife. (I’ve called this the “kickoff and cold cuts” approach to data governance, in reference to the catered lunch found in most data governance launch sessions.)
It seems that those with a stake in data governance are far more invested in the “whos” than the “hows.” That will change as executives inside and outside of IT realize that processes for defining policies and rules for data, as well as maintenance of those processes and rules, are at the crux of data governance. Companies will back up and begin by designing their data governance programs in a deliberate way, incorporating data quality, MDM, and emerging workflow tools, thus rendering decision-making bodies as only one component of a successful data governance effort.
3. There will be a renaissance in the world of business requirements. This time, however, there’s a twist. Ten years ago we were falling all over ourselves to facilitate end-user workshops, conduct structured interviews, and obtain sign-off on business requirements documents. Nowadays, business users have requirements fatigue. If someone asks them one more time what keeps them up at night, they’re going to hurl themselves out a window-likely the one overlooking your car in the parking lot. Vendors such as Serena and Kalido offer tools that capture business requirements and generate documents and models. This allows business analysts to deliver results as they’re engaging their users. Say goodbye to redundant conversations and learning curves for staff and consultants, and hello to requirements as an organic and continuous process.
4. CIOs will be asked to apply economic rigor to their IT investments. While not strictly a BI issue, this issue nevertheless has BI implications. Take storage. If a CIO can pay $200 for a terabyte at Best Buy on Sunday, how does he justify $40,000 for a terabyte on Thursday? It’s time that CIOs focus not only on total cost of ownership but cost per user and quantified business value. CIOs need to become discerning buyers seven days a week.
5. Management will realize that BI requires specialized skills. No, I’m not going to predict the rise of the BI competency center. (As much work as I’ve personally done in the area of designing BI competency centers, the truth is that the vast majority of companies haven’t yet earned the right to deploy them.) I’m going further out on a limb and predicting that executives will finally get that BI requires a new development rigor—from design reviews to spec development to detailed regression planning—and that it is simply more complex to deliver than a packaged application.
We’ll hear fewer business execs request a “data warehouse in a box,” as more realize they need to buckle down and hire experienced staff for customized, data-intensive, and atypically business-driven development. (Note to competency center acolytes: Let’s talk in 2010.)
6. Management still won’t care about metadata. Stop trying to educate them, and God forbid, don’t ask them for funding! Just bake it into your existing projects whenever you can. It’s like putting broccoli in the blender, then mixing the pasty green mess into the brownie batter. They won’t know, but it will be good for them.
7. MDM and BI will marry. For the last couple of years, BI has been eyeing MDM from afar. Meanwhile, MDM has been canoodling with business people whose urgency to streamline mergers and acquisitions or comply with complex regulatory mandates has made it the belle of the ball. Sometimes you’ll spy MDM dancing with an intrigued enterprise architect or, more brazenly, with an entire application development team. It has flirted with manufacturing and gotten serious with government, healthcare, and financial services. BI will come calling soon, its own dalliances with data quality tools and operational data stores having proved less than fulfilling.
When BI does finally call, MDM will sense a kindred spirit, a partner in all things data. Family squabbles about “analytical” and “operational” MDM will go away. Neither will quit a day job and each will keep a separate set of friends-but it’ll be a match made in heaven.
8. Identity resolution will get its due. Sure, you can call it infrastructure. Processing and rules intensive, customer identity resolution has been relegated to the underlying algorithms of third-party data providers, MDM, and data quality vendors. However, companies are recognizing that they may have unique customer data-matching needs-a bank we work with has more than 50 definitions of a household-and they’ll be looking at smarter, more specialized ways to automate them.
9. More operational software vendors will hang out the BI shingle. It’s already happening. Business application vendors, many of which couldn’t spell BI five years ago, are now offering it as core functionality. For instance, demand-chain automation vendor JDA Software offers a range of prebuilt and user-friendly reports for supply chain analysis, revenue and inventory forecasting, and other needs. Epic, claimant to a growing market for clinical software in healthcare, has built more than 100 operational reports, all fed from its transactional database.
In short, the systems that are running our companies are now also vying for a piece of our BI reporting. If you need to know the number of discharges across your hospital network in the last hour, should you load that data into your data warehouse first or simply run a report directly from the operational system? The answer: It depends. That’s the answer that more operational applications are depending on.
10. BI professionals will look beyond BI to build their resumes. In my 2008 predictions, number one was “The emergence of data as a service.” You can pat me on the back next time we see each other. Seriously, though they take different forms depending on industry and company size, data integration and governance efforts have earnestly enlisted BI folks. Why? Because BI professionals understand cross-functional data, know how it’s used, and monitor what it looks like across the data supply chain. Sometimes they’re the only people in the company that really understand data integration.
As market consolidation continues to distill the BI market into a smaller number of players, BI professionals will realize that the true value of BI might not be in the tools or the databases they use but in the data itself. When quantifying the value of software tools, databases, and data, only one reliably appreciates over time. If I were wagering my job, that’s the horse I’d bet on for 2009.
Jill Dyché is a partner and cofounder 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. She is the co-chair of TDWI’s Master Data Management Insight conference, to be held in Savannah, GA, from March 8-10. Her new blog, Inside the Biz, focuses on the business value of IT.8