Business intelligence: Take a step back

Bill Hewitt, CEO of Kalido, looks at why business intelligence systems have often failed to deliver on their promises and how a new approach can change this.


The business world is frustrated. Before computers, the accounting department was the hub of every company, recording transactions, updating ledgers and providing statements of financial condition. The computer brought much needed automation to mundane, repetitive processes, such as payroll, bookkeeping and order taking, dramatically improving productivity.

Over the past 20 years however, we’ve seen the global economy boom and information stored within IT systems become increasingly complex and difficult to manage. Data now stored in a myriad of systems using wildly different methods drove the IT community to look for a solution beyond traditional reporting. Today, companies have spent (and carry on spending) vast sums on the supposed aspirin to this data headache - business intelligence (BI).

The increasing need for business managers to make faster, more strategic decisions based on the plethora of information delivered to them means that prioritization and organization of data has become a huge priority for many businesses. Further, the business insight that was important during economic boom times can yield the competitive advantage needed to survive during a worldwide recession.

With such heavy investment - Gartner predicts the BI market to hit $7.7 billion market in 2012 - users’ expectations are justifiably high. According to Gartner, organisations implementing BI principally expect it to speed up and improve organisational decision making, enabling them to quickly respond to user needs for data and better align and track against corporate strategy and objectives.

This comes in addition to the cost savings and efficiency improvements anticipated with most IT solutions. However, when asked about BI, respondents to an Economist Intelligence Unit report revealed that a staggering 72 per cent of senior and IT business leaders said their data was still inconsistent across their departments, while 40 per cent said their workers made poor decisions because of it. It’s therefore little wonder they are so annoyed.

So why has BI failed so spectacularly to live up to the hype? For this, we need to look at the lifecycle of a typical company. As a company grows, it begins to form separate departments dedicated to particular parts of the business, often significantly increasing the business’ complexity and creating silos of data - and information.

Separate business units may appear, usually operating autonomously - each making its own decisions and each with its own IT department. At some stage, information from these units are often brought back together to get a ‘corporate’ view - or the company is centralized entirely - but the mass of disparate information sources leads to major reporting problems.

Even basic reporting, on a weekly or monthly basis, becomes a struggle. So the problem does not lie with the BI tools themselves, but rather what’s behind the BI - the outdated processes and systems used to marshal the information and service it up to the BI tools for delivery to the end users.

The simple fact of the matter is that these environments use IT tools designed to help build the environment efficiently, such as data modelling tools, ETL and data quality products, all working on their own small part of the data. Having a toolbox is great, but trying to use a hammer, screwdriver, saw, level and pliers to make a car run is a highly inefficient proposition.

"Recommended For You"

Oracle makes BI play nicely Everything you wanted to know but were afraid to ask about BI