InsightaaS: We have seen a surge of interest in analytics as organizations come to appreciate the power of applying data-driven analysis to business decisions. A recent article in the McKinsey Quarterly notes that analytics supports many critical operations in different industries: customer-facing activities (for example, in helping telecommunications companies with promotions and retention), internal applications (where analytics increase the efficiency of processes like route optimization and maintenance scheduling) and hybrid applications (such as in retail, where analytics provides insight into both buyers and supply chain). The piece notes that “Big-data analytics are delivering an economic impact in the organization,” but adds that “too often, senior leaders’ hopes for benefits are divorced from the realities of frontline application.”
How can IT and business management align the potential of analytics with its real-world benefits? The article points to two potential adoption paths: training, which enables specialists to take advantage of sophisticated tools, and automation, in which analytics is embedded in “intuitive end-user interfaces that can be rolled out rapidly and with little training.” It seems clear to us that the former path will be applicable in specific contexts, but that the latter will be the primary means of distributing analytics benefits broadly across and within public and private-sector organizations.
It is clear that there is widespread interest in creating intuitive end-user interfaces embedding analytics output. At a recent full-to-capacity “Visualize Responsibly” event in Toronto, Chris Banks,Â director of BI and performance management solutions for Information Builders, discussed the key considerations in building analytics-based rich-interface solutions that are widely used — and have widespread benefit — as means of enhancing business processes.
One important part of a virtualization assessment is understanding the reasons why a visual approach to exploring and analyzing information is important. According to Banks, “we’re all very visual people — half of [our] brains are absorbing information…we see things very quickly, very easily, we retain information, we start doing recognition” from visual inputs. Visual connections are faster and more intuitive than connections based on other presentation modes — and “faster and more intuitive” are essential elements in capturing business benefit from many types of technology, and particularly, from advanced solutions like analytics. Banks emphasized this point through his visualization presentation: “around BI,” he said, the “philosophy is simple: make [data] easy to use. Regardless of whose technology you have, if you make it easy to use, I promise you people will use it.” This is the key driver of data visualization — “when you have large amounts of information…and look at it visually, you can start doing analysis more quickly and easily. ”
Banks offered several examples of making BI simple and flexible, driving use of analytics to executives/management and other non-technical business users. The first started with Excel, which Banks positioned as the most popular analytical tool in use today; he demonstrated that the key to ensuring “one single version of the truth” is to connect (as Information Builders’ tools do) spreadsheets to source data, drill-down paths and calculations, so that all recipients of an Excel-based model have a common view of what the model is based on and represents.
The second approach demonstrated in Banks’ presentation was a guided self-service report. His example included build-once, maintain-once reports that offer millions of possible output combinations, based on different data parameters and output formats. This type of structurally-sound approach to analytics offers advanced users the flexibility they require, but reduces the need for custom programming (by either IT or the users), and results in making the back-end data “transparent to the user” — managed, securely and through a single system, by IT.
Neither or these approaches, or other paths to effective enterprise BI, is as simple as a first-time user might anticipate. There are several challenges that must be addressed in order to ensure that visually-rich, analytics-based applications deliver the value that early adopter organizations (and their senior management) anticipate.
One critical issue is data quality. Banks quoted author Larry English as saying that organizations are spending 20%-35% of their operating revenue due to data quality issues, and cited a second source — The Data Warehousing Institute (TDWI) — whose statistics hold that data quality problems cost U.S. businesses more than $600 billion a year. These figures are staggering in aggregate, but to some extent, they are easy to ignore in specific business contexts because the inconsistencies driving these costs are scattered throughout the organization and hidden in multiple business silos and tools. With an enterprise visualization strategy, though, data quality issues are exposed: management across the enterprise gains visibility to information that is at odds with their experience and/or with other facts that they can access and compare. Data integrity (ensuring that the structure of the data corresponds to usage requirements) and effective master data management (the definitions and processes used to align data with its use across different contexts) also pose challenges to data-dependent reporting, especially when organizations attempt to connect internal records with external data sources.
While analytics dashboards expose these data issues, the lack of advanced analytics tools also exacerbates the problem. Organizations that are managed via multiple Excel worksheets will find that they have “multiple versions of the truth,” or different spreadsheets containing different values for common factors such as sales, inventory, pipeline value, etc. From a data quality perspective, there are two related keys to effective visualization: the ability to draw upon credible data, and the ability to draw consistently upon that data, so that a single variable (sales, inventory, etc.) is both factually correct and presented according to common understandings wherever it is used.
The presentation also helped attendees to understand the functionality and limitations associated with different sub-categories of BI. For example, data discovery is helpful in identifying problems, provided that data quality (including data integrity and master data management) is sound. However, data discovery tools don’t address some of the outcomes that executives are coming to expect from BI. To illustrate how a fully-functional BI system expands on the scope of a data discovery tool, Banks invoked the analogy of the Wonkavator — the mythical elevator capable of going “sideways, and slantways, and longways, and backways…and squareways, and front ways, and any other ways that you can think of,” found in the movie Willie Wonka & the Chocolate Factory. In the analogy, a data discovery tool is an elevator, excellent for use within a specific set of requirementsÂ — but “it’s not good at what it doesn’t do [such as] predictive analytics…enterprise search, solution analytics, portals, mash-ups.” Banks asked, “How do I integrate all that together?” adding “I don’t want business users running amok. How do I manage them, and secure them, and allow them the drag-and-drop that they want to see?” At a fundamental level, this leads users to make a choice in technology selection. Are they looking for pre-built applications that will deliver answers to users without requiring training or in-depth understanding of the underlying data? Are they looking for a tool which provides flexibility, but requires that users be trained? Or are they looking for a platform capable of connecting tools and applications into a broader suite of BI capabilities?
The conclusion of Banks’s speech provided guidance to attendees looking to parse key issues in BI adoption. “Every organization is different,” he observed, and as a result, “there is no one right approach; you can go down the tools road, or you can use the app approach,” managing the business users by allowing them to select pre-defined views of data from a virtual app store. The key, Banks said, is to “get BI into the hands of everyone in your organization, and get them to use it.”
Summary: In the end, Information Builders is a platform vendor with an approach that is clearly on the ‘automation’ side of McKinsey’s automation vs. training debate. By enabling use of data through visually-rich interfaces embedded within process applications, and by supporting more advanced users through guided self-service portals, Information Builders is creating a path to widespread, high-value use of analytics within the enterprise.