What if business intelligence really delivered intelligence?

Businesses talk a lot about data-driven decision making. But it will take a new approach to business intelligence to really achieve it.

Data-driven decision making and business intelligence systems go hand-in-hand. But while the typical BI dashboard includes lots of data, there’s very little context for it, and it’s not targeted at individual decision makers.

A sales dashboard, for instance, might reveal that a competitor’s sales surged last month. But to figure out why, decision makers would need to consult an array of systems—from summary reports on product sales performance, to external research on market conditions—to help them decide what to do in response.

Or consider a marketing leader with separate business intelligence (BI) reports on campaign strategies, strategic initiatives, marketing plans and tactics but no way to easily piece everything together to understand campaign effectiveness, conversion rates and uplift. Then there’s the business unit leader with hundreds of data points from the supply chain, manufacturing, sales and marketing functions but no way to synthesize it all and figure out what changes are needed to improve.

Bottom line: it takes a long time to sift through all the data to really figure out what’s going on.

The fact is, every individual business user takes a different and dynamic path to arrive at an insight and decide what the best action is to take. And BI systems need to accommodate these individualized analytic pathways to truly drive data-based decision making.

One idea is to take an approach similar to how Netflix serves up movies and movie suggestions from an array of sources, based on what the algorithm behind it learns about individual behaviors and preferences.

A Netflix approach to data-driven decision making

To understand individualized analytic pathways, consider what happens when your car engine light turns on. One driver might suspect it’s the battery. Another might figure the oil needs changing. Each will want additional (but different) data. 

The first driver might want to know how long batteries typically last. The second might want to look up when the last service appointment was. Each will need to seek outside information sources, and based on the information they find, make a decision: take it to the dealer, a battery supplier, the local service station or ignore it and hope the warning light eventually goes away. 

What if they could get that information from the car dashboard? Or even better, what if that additional data was served up to them based on what they’ve already stated were their information preferences, how they’ve behaved over time and even the data that others like them were looking for in a similar situation?

Such a dashboard would be similar to a Netflix experience: Information is clearly laid out based on established preferences, it dynamically changes based on actions the subscriber takes, and it serves up suggestions based on what similar-minded viewer are interested in. The dashboard would also provide a drill-down capability so that users can see additional information that would help contextualize and provide more in-depth insights into what’s behind the data being displayed.

Data-driven business intelligence in the workplace

Here’s what this might look like in a business setting. Typically, sales executives are faced with overwhelming amounts of data—sales figures, customer data and market trends among others. Extracting meaningful insights is time-consuming and challenging, with the typical dashboard only showing disparate statistics like total sales, sales by geography, comparisons with competitor sales, etc., without explaining why certain trends occurred or recommending actionable decisions.

The sales executive would have to spend a significant amount of time going through various reports to piece together the story and generate insights for further decisions.

With a Netflix-like dashboard, the sales exec would also get context alongside the data. For example, if sales surged, the dashboard would explain the contributing factors, like a successful marketing campaign or a new product launch. Just as Netflix suggests a movie, the dashboard would also recommend actions. If competitor sales rise, it would propose strategies, such as targeted promotions or a competitive analysis.

Further, the more the sales executive interacts with the dashboard, it would dynamically change and adapt itself to the persona and provide more contextual information. With the dashboard’s drill-down capability, the sales executive could explore deeper layers of information, such as analyzing individual customer behavior or regional performance.

Putting the intelligence into business intelligence

To create this type of system, businesses would need a system that was able to do the following: 

• Pull data from the multiple BI systems they have in the enterprise, along with a wide array of additional data sources, both inside and outside the business
  
  
• Provide a searchable catalog of regularly refreshed content, accessible via natural language search
  
  
• Create an easy-to-navigate display that’s as intuitive as what you’d see in a consumer setting
  
  
• Provide drill-down capabilities that provide additional context for the data, pulled from other data sources
  
  
• Enable users to tell “stories” by compiling data points into a logical flow that explains why a particular situation exists
  
  
• Regularly remove insights that the individual user doesn’t tend to use
  
  
• Use AI analytics to provide alerts, notifications and recommendations

We worked with a life sciences company to create a business intelligence system that operates in this way. Senior executives at the organization were struggling to get a comprehensive view of the business, across payers, pharmacies, hospitals and field sales. Throughout the organization, the existing business intelligence and reporting applications were underutilized, which led to lots of time spent mulling over reports manually.

The reports were not personalized to the needs of each decision maker, and critical metrics remained unidentified, hindering the identification of causal relationships.

We worked with the company to implement a business intelligence platform that offers hyper-personalized user experiences, a searchable content catalog, content recommendations, standard interface design and efficient content deployment.

Now, the company has a unified platform for business insights across functions that consolidates 60 disparate data sources. Managers spend far less time generating reports and can much more quickly identify KPI-related causal relationships. The platform has also reduced annual costs by $1.2M for legacy operational dashboards and global management reports.

Time to redefine data-driven decision making

Business intelligence and data-driven decision making have long been sought after—and long been criticized for delivering less than expected. It’s time for a new way, one that borrows from a familiar, popular experience and responds to how people actually think, decide and act.