The marketing and analyst airwaves are flooding with speculation about what is next for business intelligence (BI). What will comprise BI 2.0?
Historically, this market has been served by vendors such as Business Objects and Cognos. But the competitive landscape is changing. Microsoft has now shrewdly entered the market by driving the placement of SQL servers into the space in order to broadly deploy and deliver its BI suite and reporting services in volume. Oracle has seen the effect of companies moving data out of the database to stage it for analysis. The resulting data warehouses have provided a degree of utility in housing, manipulating and delivering “strategic” information across the organization.
Recently though, established vendors such as SAP and Siebel have unveiled BI product suites under the banner of “analytics.” SAS, a perennial stalwart of the statistics market, is suddenly being touted as the number-three BI vendor and frequently positions itself as an analytics vendor.
With analytics finding its place within many functions and business processes it seems clear that it will be a defining feature of next generation business intelligence. Particularly, a significant new group of business users—a group I like to call “Go-To Guys”—are in need of analytics tools to tackle daily problems and opportunities. Go-To Guys are the operating managers of company—product managers, sales managers, researchers, engineers and marketers.
So, what is analytics? Neil Raden of Hired Brains, a market research and management consulting firm, has said that, “the proper term for interacting with information at the speed of business, analyzing and discovering and following through with the appropriate action, is ‘analytics’.”
CIOs often assume that business analytics (BA) comes along with BI. The traditional BI market has been associated with providing executive dashboards and reporting to monitor the assumptions and key performance metrics that are part of long term planning cycles.
Everybody wants a dashboard. To the extent that all of us are CEO’s of our own business discipline, we want a simple measurement display of how we are doing and an alert mechanism of when something goes wrong. Additionally, dashboards address the growing urgency around Sarbanes Oxley. Monitoring planning assumptions and key performance metrics has now become mission critical from a regulatory and compliance standpoint.
Where BI Stops and BA Begins
But BI reporting ends with the dashboard, which is sufficient only for some business planning, and BA picks up the rest for the Go-To Guys. Simply, this group must interact with data in a much different way from what traditional BI allows.
The Go-To Guys deal daily in unanticipated outcomes and unknown results and it is their job to mitigate risk and capitalize on opportunities. BI is not architected to iterate on new scenarios or for immediate response to unanticipated questions because it is set up to automate the distribution of standardized reports that monitor pre-determined key performance metrics and planning assumptions. BI’s answer to analytics has been to deliver the report to the business user and the business user typically takes the data in the report and dumps it into Microsoft Excel in order to do his own analysis.
As a result, there are $8B (yes, billion) of internally developed analytic applications with Excel as their front end. The BI players treat
the output to Excel as a feature. But I actually think it’s a tremendous failing. It is proof that you don’t get BA when you buy BI. The BI architecture cannot support the operating needs of the business users to ask and answer their own questions in response to new occurrences and events in the marketplace.
Secondly, Excel is not an answer either. As soon as the data is dumped into Excel, the user is out of the BI system with no way back in. Any insight that the business user gains while interpreting Excel spreadsheets tends to stay with him—all opportunity for organizational learning or process improvement is lost.
So requirements for analytics are different than the requirements for BI, but the benefits are different as well.
There’s also a technical component to all of this reinforcing the claim that the technical requirements to support analytics are different from the technical requirements that enable BI. To facilitate reporting and dashboards, BI traditionally works with aggregated data. Business users cannot rely solely on aggregated data in the operating environment. They have to be able to get to the details. The aggregated data will many times obscure the key issue or opportunity in your information.
BI data is typically staged in an OLAP cube to support drill-down. In analytics the Go-To Guys have to be able to get directly to the source data in the database. The key facts needed to make your operating decision are often not in the cube because they haven’t been anticipated by the IT department. This is not a question of the trees obscuring the forest—you have to be able to see both. The business users cannot be disconnected from the critical data needed to make a key business decision.
And lastly, the requirement of the BI system has been to monitor the data based on pre-configured questions requiring only a thin client environment to inform the user. In the operating world, users need to engage with the information requiring a richer client to support interactivity and the ability to ask and answer their own question without having to go back to IT.
What are the characteristics of an analytic savvy organization? First of all, even the planners want into the act. Analytics is enabling more proactive, high-frequency planning cycles. Planners are better able to refine and iterate the plan, shifting resources to higher performing areas with the goal of being first-to-market and never having a warehouse full of trendy goods once the trend is over. Secondly, the analytically savvy organization is more agile—able to adapt and respond—whether that’s to a competitor that releases a new product, a change to the pricing structure in the marketplace or the success of its own marketing campaign.
Remember, you don’t get business analytics when you buy business intelligence. The requirements are different and the benefits are different. The return on information and expertise achieved by arming your operating managers with analytics will supercharge your existing BI investment.
Rock Gnatovich is the president of Spotfire, Inc. , a provider of analytics software. Rock has spent 18 years in the software industry, with the last 10 focused on enterprise applications. You can reach him at firstname.lastname@example.org.
Follow everything from CIO