Get Smart! 3 steps to better business intelligence

Businesses today are continually asking questions about how to cut expenses and improve profitability. Increasingly, a comprehensive Business Intelligence solution is the tool they’re turning to for answers – David McNickel finds out how it’s done...

 

So what exactly is Business Intelligence (BI)? Although the concept of BI can be fairly daunting – conjuring up images of giant computer banks and legions of tech-staff – in reality it can be summed up in a sentence. Business Intelligence is nothing more than the process of increasing a business’ competitive advantage by the intelligent use of available data in decision making.

September 2003

To use a medical analogy, if information is the corporate lifeblood – then you need to know how healthy you are and weigh up your vital signs against your competitors. Like many ebusiness related solutions, the implementation and technology costs of BI have meant it has been the exclusive domain of very big businesses in the past. More recently, however, software costs and development times have come down significantly – making BI an affordable option for many.                     

So what does it offer? The benefits of BI may include identifying top customers, product line profitability, demographic trends and fine-tuning of pricing policies, in addition to customer retention and predicting market trends.

A few examples of local organisations using BI successfully are:

  • Progressive Enterprises – using its Foodtown and Woolworths onecard and a Business Intelligence solution to analyse real-time customer data and run more effective marketing campaigns.
  • Christchurch International Airport – using Business Intelligence to analyse aircraft traffic, passengers and trends to get the most from its marketing budget.
  • The University of Waikato – using Business Intelligence to reduce report-processing time – in many cases from days to minutes.
  • Fisher & Paykel Finance – using Business Intelligence to identify customer buying patterns for future finance agreements, and to help plan staffing arrangements for customer service programmes during peak periods. In general terms the BI process is made up of three components or ‘steps’.

Step 1: Collect the data

Technological advances have made it possible for businesses to gather and store huge amounts of data about their day to day activities.

Flowing in from numerous sources, this data typically includes customer contact, financial information, operations and transactional data. In order to be any use at all in a BI sense, however, it must all be captured electronically – warehouses full of paper files are of little use.

Typically, when approaching a BI solution, most companies find that over the years they’ve built up a number of disparate legacy systems and databases – some Oracle here, some IBM there. Routinely these systems aren’t integrated and they may be difficult to extract information from. It is at this stage of the process that it becomes clear that ‘data’ and ‘information’ are two different things – and data on its own is not enough to make decisions.

The good news is that most modern BI software can make use of legacy information stored in these disparate systems, meaning old databases don’t necessarily have to be replaced. The technical term for this process is to perform an ETL function – which is to Extract, Transform and Load data. On a continuing basis, BI applications extract data from databases or transaction systems, check it for errors, clean it up and translate it into a uniform format.

Step 2: Data analysis

This is the crunch stage of a Business Intelligence solution – the raw data is coal – and you’re looking for diamonds. Once again, although the thought of getting useable Business Intelligence from a collection of disparate databases and applications may seem daunting, the process is actually relatively easy. The first step to take in the Data Analysis stage, is to decide exactly what it is you (and your staff) need to know. Essentially this boils down to what are the key business decision needs for your company.

What questions do managers need answered – and how do they know whether performance is improving or declining? Of course for different department heads the informational needs will be different. Sales people want sales and marketing info, production people want volume estimates, while customer service managers want the facts about their company’s relationship with its customers.

A BI system needs to filter data into information that’s relevant to decision makers and it should also present the information in a way that’s easy for them to review. Many BI systems allow users to view data in a variety of different formats including tables, charts and graphs.

These tools help knowledge workers understand and use information in ways they might never have previously considered – including the creation of many ‘what if?’ and best-case/worst-case scenarios. But that’s only if they use them. Research has shown that everybody absorbs information differently. Some people learn best with graphics, while others respond to text or numbers. You need to understand how your knowledge workers like their info presented. A BI solution that nobody likes using could potentially be a huge waste of money.

The basics

As a starting point, a Business Intelligence solution should provide the following:

1.

Fundamental analysis tools so staff can do what they already do with existing spreadsheets or other methods.

2.

Insight into your company’s performance that would be difficult or impossible with your current tools.

3.

Analysis that goes beyond the surface, so staff can look behind the numbers to get to the source of a problem.

4.

Information delivered in context so that any reasonably intelligent person could make an informed decision.

5.

The facility to produce reports automatically so that information can reach people when they need it.

Step 3: Take action

Business Intelligence has no intrinsic value until it is used. In simple terms, knowledge workers using their organisation’s BI systems must feel empowered to make decisions – and take action accordingly. In many respects this step of the process should be carefully considered at the outset of BI development, as the only truly successful BI implementation is one with complete senior management buy in.

What should be included in a Business Intelligence budget?

The obvious first thing is the purchase cost. You will pay seat licences for everybody who’ll use any piece of the solution. Be careful not to overbuy here. Match the needs of each person with the most appropriate tool (80% of the people using your system need only view reports where they can change a few parameters – this is generally the least expensive component of the software).

Realistically, you don’t want everybody performing ad-hoc queries all day long. It will overwhelm your servers and your network (adding a whole new expense to your budget). Too much functionality can be as big a mistake as not enough. It adds complexity. And complexity leads to headaches – headaches you might pay extra to enjoy.

Next, budget for implementation costs. Fast talkers will gloss over the many big hidden implementation costs. Use caution here. Don’t pay for things you don’t need. Don’t pay for systems that replicate your entire environment, because then you’ll pay to have someone recreate the redundant system every time you upgrade, change, or add data sources – compounding the amount of double work and costs.

Then there are training costs for all of those people (depending on the solution you choose). No matter what you choose, plan for some kind of training. Even if it’s training you develop and deploy yourself. Finally there are maintenance costs. This is the part most vendors leave out of their ROI discussions.

Typically, it runs between a third and a half of your total cost. And you’ll pay that much every year. Just accept that new data sources will be added on a regular basis. Then accept that your IT staff will need to modify the metadata (at least) to accommodate the new information supply. And budget for these maintenance costs.

These are the obvious costs associated with a typical business intelligence solution.

(Taken from Brio’s Intelligent Guide to Business Intelligence)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

For more information on Business Intelligence visit the leading providers of BI tools and solutions in the Business Intelligence  Research Pavilion


 

site by doubleclique