Business intelligence (BI) is an evergreen that simply refuses to give up and get commoditized. Even though very few vendors try to differentiate these days on commodity features like point and click, drag and drop, report grouping, ranking, and sorting filtering (for those that still do: Get with the program!), there are still plenty of innovative and differentiated features to master. We categorize these capabilities under the aegis of Forrester agile BI; they include:
Making BI more automated: suggestive BI, automatic information discovery, contextual BI, integrated and full BI life cycle, BI on BI.
Making BI more pervasive: embedding BI within applications and processes, within the information workplace, and collaborative, self-service, mobile, and cloud-based BI.
Making BI more unified: unifying structured data and unstructured content, batch and streaming BI, historical and predictive, and handling complex nonrelational data structures.
Too little data, too much data, inaccessible data, reports and dashboard that take too long to produce and often aren’t fit for purpose, analytics tools that can only be used by a handful of trained specialists – the list of complaints about business intelligence (BI) delivery is long, and IT is often seen as part of the problem. At the same time, BI has been a top implementation priority for organizations for a number of years now, as firms clearly recognize the value of data and analytics when it comes to improving decisions and outcomes.
So what can you do to make sure that your BI initiative doesn't end up on the scrap heap of failed projects? Seeking answers to this question isn't unique to BI projects — but there is an added sense of urgency in the BI context, given that BI-related endeavors are typically difficult to get off the ground, and there are horror stories aplenty of big-ticket BI investments that haven’t yielded the desired benefit.
In a recent research project, we set out to discover what sets apart successful BI projects from those that struggle. The best practices we identified may seem obvious, but they are what differentiates those whose BI projects fail to meet business needs (or fail altogether) from those whose projects are successful. Overall, it’s about finding the right balance between business and IT when it comes to responsibilities and tasks – neither party can go it alone. The six key best practices are:
· Put the business into business intelligence.
· Be agile, and aim to deliver self-service.
· Establish a solid foundation for your data as well your BI initiative.
Market conditions are changing quickly; firms need to make the best possible business decisions at the right time and base them on timely, accurate, and relevant information from business intelligence (BI) solutions. The repercussions of not handling BI change well are especially painful and may include lost revenue, lower staff morale and productivity, continued proliferation of shadow IT BI applications, and unwanted employee departures. Ineffective change management often lies in the process of preparing the people affected by change rather than in planning the technology implementation. Firms that fail to prepare employees for enterprise BI change early enough or well enough will be left behind. They need to implement a multifaceted series of activities ranging from management communication about why change is needed to in-depth, role-appropriate employee training.
Why change management is so critical? Most strategic business events, like mergers, are high-risk initiatives involving major changes over two or more years; others, such as restructuring, must be implemented in six months. In the case of BI, some changes might need to happen within a few weeks or even days. All changes will lead to either achieving or failing to achieve a business result. There are seven major categories of business and organizational change:
Business process changes
New technology implementations
Changes to business process outsourcing or IT sourcing
Initial business intelligence (BI) ployment efforts are often difficult to predict and may dwarf the investment you made in BI platform software. The effort and costs associated with professional services, whether you use internal staff or hire contractors, depend not only on the complexity of business requirements like metrics, measures, reports, dashboards, and alerts, but also on the number of data sources you are integrating, the complexity of your data integration processes, and logical and physical data modeling. At the very least Forrester recommends considering the following components and their complexity to estimate development, system integration and deployment effort:
The Obama 2012 campaign famously used big data predictive analytics to influence individual voters. They hired more than 50 analytics experts, including data scientists, to predict which voters will be positively persuaded by political campaign contact such as a call, door knock, flyer, or TV ad. Uplift modeling (aka persuasion modeling) is one of the hottest forms of predictive analytics, for obvious reasons — most organizations wish to persuade people to to do something such as buy! In this special episode of Forrester TechnoPolitics, Mike interviews Eric Siegel, Ph.D., author of Predictive Analytics, to find out: 1) What exactly is uplift modeling? and 2) How did the Obama 2012 campaign use it to persuade voters? (< 4 minutes)
In advance of next week’s Forrester’s European Business Technology Forums in London on June 10 and 11, we had an opportunity to speak with Greg Swimer about information management and how Unilever delivers real-time data to its employees. Greg Swimer is a global IT leader at Unilever, responsible for delivering new information management, business intelligence, reporting, consolidation, analytics, and master data solutions to more than 20,000 users across all of Unilever’s businesses globally.
1) What are the two forces you and the Unilever team are balancing with your “Data At Your Fingertips” vision?
Putting the data at Unilever’s fingertips means working on two complementary aspects of information management. One aspect is to build an analytics powerhouse with the capacity to handle big data, providing users with the technological power to analyse that data in order to gain greater insight and drive better decision-making. The other aspect is the importance of simplifying and standardizing that data so that it’s accessible enough to understand and act upon. We want to create a simplified landscape, one that allows better decisions, in real time, where there is a common language and a great experience for users.
2) What keys to success have you uncovered in your efforts?
As an analyst on Forrester's Customer Insight's team, I spend a lot of time counseling clients on best-practice customer data usage strategies. And if there's one thing I've learned, it's that there is no such thing as a 360-degree view of the customer.
Here's the cold, hard truth: you can't possibly expect to know your customer, no matter how much data you have, if all of that data 1) is about her transactions with YOU and you 2) is hoarded away from your partners. And this isn't just about customer data either -- it's about product data, operational data, and even cultural-environmental data. As our customers become more sophisticated and collaborative with each other ("perpetually connected"), so organizations must do the same. That means sharing data, creating collaborative insight, and becoming willing participants in open data marketplaces.
Now, why should you care? Isn't it kind of risky to share your hard-won data? And isn't the data you have enough to delight your customers today? Sure, it might be. But I'd put money on the fact that it won't be for long, because digital disruptors are out there shaking up the foundations of insight and analytics, customer experience, and process improvement in big ways. Let me give you a couple of examples:
BI professionals spend a significant portion of their time trying to instill the discipline of datadriven performance management into their business partners. However, isn’t there something wrong with teaching someone else to fly when you’re still learning to walk? Few BI pros have a way to measure their BI performance quantitatively (46% do not measure BI performance efficiencies and 55% do not measure effectiveness). Everyone collects statistics on the database and BI application server performance, and many conduct periodic surveys to gauge business users’ level of satisfaction. But how do you really know if you have a high-performing, widely used, popular BI environment? For example, you should know BI performance
Efficiency metrics such as number of times a report is used or a number of duplicate/similar reports, etc
Effectiveness metrics such as average number of clicks to find a report and clicks within a report to find an answer to a question and many others
Metric attributes/dimensions such as users, roles, departments, LOBs, regions and others
Whether you are just starting on your BI journey or are continuing to improve on past successes, a shortage of skilled and experienced BI resources is going to be one of your top challenges. You are definitely not alone in this quest. Here are some scary statistics:
“By 2018, the United States alone could face a shortage of 140,000 to 190,000 people with deep analytical skills, as well as 1.5 million managers and analysts with the know-how to use the analysis of big data to make effective decisions.” (Source: May 2012 McKinsey Global Institute report on Big Data)
“… trigger a talent shortage, with up to 190,000 skilled professionals needed to cope with demand in the US alone over the next five years.” (Source: 2012 Deloitte report on technology trends)
“Fewer than 25% of the survey respondents worldwide said they have the skills and resources to analyze unstructured data, such as text, voice, and sensor data.” (Source: 2012 research report by IBM and the Saïd Business School at the University of Oxford)