Over the past decade, organizations have built data warehouses, purchased BI tools, supported adoption by end users and applied the information to make business decisions. Many have mastered the art of turning raw data into some kind of information or actionable insight.
That does not guarantee that information has been used pervasively or in a timely fashion. Do your nuggets of information simply measure the business performance by looking at historical data, or do they have predictive capabilities that support proactive decisions? Do they help guide users in their day-to-day activities and suggest possible scenarios in their decision-making?
The answers to these questions, in most cases, reveal that companies are not taking full advantage of predictive analytics capabilities. It is true that through data warehouses and reporting interfaces, organizations are able to turn raw data into information. They can define a single view of their business through disciplined data definitions, taxonomies and governance programs. The technology evolution coupled with sophisticated BI applications helps users generate interactive reports or views of dimensional data, ideally leading to a point of action - a decision, a plan and/or a request for more information.
What all this work might still fail to do is help with actual decision-making in a timely manner, leaving us unable to answer the questions posed by senior executives. Unaddressed are such questions as: How do we cut down on event-to-decision latency? How do we automate common analysis tasks? How do we ensure consistent analysis? How do we capture and reuse expert knowledge? How do we get new people up and running quickly?
While sophisticated BI systems provide a maturity roadmap starting from data collection and standard reporting capabilities to analyze past performances, predictive analytics focuses on modeling to create competitive advantages. These separate objectives inherently create gaps in an organization's analytics maturity.
The gulf - moving from report browsing to finding patterns, correlation between events and subsequent implications, etc. - needs to be bridged with navigational guidance, automated alerts and what-if scenarios. This can be approached by spending more time analyzing data to make fact-based decisions and optimizing day-to-day operational activities.
Bridging the chasm is about leveraging information as an enterprise asset and predicting future outcomes through deep data analysis, modeling, forecasting and simulations. This is a huge leap from reviewing reports and making reactive decisions.
Guided analysis provides timeliness in dealing with anomalies and allowing the workforce to spend time performing quality work and managing greater workloads without any reduction in efficiency. If the effort in analyzing data to see if a problem is present can be reduced, more effort can be expended to remedy problems that have already been identified.
From an implementation perspective, at its simplest level, within the reports and dashboards it's possible to customize the display layout and content relevance to users based on their profile, and to define alerts using a conditional request that will send reports to users when certain business events have occurred.
But we can do more with guided analytics. In the case of a customer calling a service representative, for example, guided analytics can add a voiceover into the operator's headset relating relevant information about the customer's past purchasing habits, recent complaints and lifestyle status to assist in prioritizing the representative's attention on the information that matters most.
With guided analytics, the realm in which predefined analytics executables are embedded into business processes, reports and dashboards becomes smarter and increasingly helps direct the decision-making process. Guided analytics will heavily influence collaborative decision-making by not only presenting relevant information, but also suggesting how the user should react to it. This may be presented as a list of ranked options, together with their respective pros and cons.
In addition, guided analytics can develop decision-making by providing the following mechanisms:
Storyboarding captures the decision-making process and best analysis paths, thereby allowing users to record these activities for future reference, collaboration and training.
"See also" suggestions provide recommendations for linked key performance indicator reports while ensuring the right decisions are taken through comprehensive analysis and sharing of best practices.
Dynamic search and auto suggest are capabilities to interpret reports and ensure users explore beyond the data to compare and analyze various implications.
Sticky notes enable users to capture thoughts as they perform a chain of analysis to make sure nothing is missed.
Personalization provides prebuilt, role-based analytic scenarios that are easily customizable, thereby making the business analysis process a time-efficient and productive endeavor.
Most organizations have a long way to go before they will be able to reap the full benefits that are now possible through predictive analytics. Technology exists to take enterprises beyond the realm of BI or descriptive analysis to use navigational guidance that will enable effective analysis of data in order to make savvy and timely decisions.