Freddy Duarte

They clearly forget to mention that Credit Unions provide loans at way better rates, and their fee schedule is so ridiculously lower.

Read Full Article

Who uses BI successfully?

• Western Digital uses BI to better manage its inventory, supply chains, product lifecycles, and customer relationships. BI has enabled the company to reduce operating costs by 50%.
• Capital One uses BI to analyze and improve profitability of its product lines as well as effectiveness of its business processes and marketing programs.
• Continental Airlines invested $30 million in BI to improve its business processes and customer service. Continental says it has reaped a $500 million return.
• A Recent Accenture study that found that nine in 10 senior executives at Fortune 1000 companies place strong analytical and business intelligence capabilities at the top of their list in preparing them for their biggest challenge ahead.

So, what is Business Intelligence?

Well, it’s a collection of tools. More like a sub-layer, actually. A sub-layer that is part of a major layer of Decision Support Systems (see diagram below). Business Intelligence sits on top of an organization’s data layer and tries to manipulate and transform that data into information. What’s the difference? Information is valuable, data is not (read more…). The Business Intelligence layer’s purpose is to squeeze out has much value as possible from your static and boring data layer, and use this valuable information to accomplish three goals:

1. Increase the organization’s profit
2. Reduce the organization’s costs
3. Improve business processes

Evolution

Let’s consider the evolution of data systems, just to put things in perspective.

1. Transaction Systems
2. Management Information Systems
3. Decision Support Systems

If you’d like more detail about this, read my previous post about storage systems.

Anyhow, we started out by static/boring transactional data systems that held basic information, like: Customer A bought product B for ABC cost. Boring, huh? A couple of years later, someone had a the great idea of creating reports from that data, and hence the MIS era was underway. The reports were great, but managers wanted more value added to their information. That’s when Business Intelligence came along. Managers wanted more information to compete better and cheaper.

The questions being asked now (and we’re talking about the 1970’s or something) were: What happens to product sales if we decrease price? What happens to our performance levels if we layoff 5 employees? What if he hire 2? How can I produce more? How can I produce faster? How can I maximize profits?

That’s the whole purpose of Decision Support Systems. It puts your business processes on rails. You have a question, a well built business intelligence layer will most likely provide you some sort of answer to make a good solid decision!

What does BI include?

• Data Warehouses, data marts
• Reporting, querying
• Dashboards
• Forecasting, statistical analysis
• Simulation, optimization models
• Business Process Management (BPM, YAWL, etc)
• Process re engineering

BI Drivers

Anything from ERP systems to web technologies, analytical software, network infrastructure and mature data warehouse technologies.

Failures

Many BI initiatives have failed to live up to their hype. A recent survey in the UK found that 87% of BI projects don’t live up to expectations. Nearly a quarter of BI projects intended to improve management decision making are going over budget. A fifth found that data failed to reveal important information, and only half said that end-users were satisfied with the system.

Reasons of Failures

Failure doesn’t happen because of technical issues, altough data quality and integration need often more attention. In general, failure happens because the business value aspect is not built into the project from the beggining and many projects aren’t “smoothly” shown to users, hence making them quit when first obstacle is found.

My main source for this post is the book we are using in the Decision Support Systems class I’m taking this semester: Quantitative Analysis for Management 10^th Edition.

The authors of the book don’t mention aspects that seem obvious when making a “good” decision. For instance, wouldn’t you think that in order to make a good decision one would need to have some level of cognitive ability, or previous experience, to deal with the problem at hand? Moreover, maybe your decision will be largely based on cultural or religious beliefs, rather than what’s really the best solution. In other words, the authors decide to take very pragmatic and mathematical approach to decision making.

So, whether you’re thinking about getting a hair cut today, voting republican or democrat, or laying off 1000+ employees, here are the list steps 6 steps of the decision making process.

Six Steps in Decision Making

1. Clearly define the problem at hand.
2. List the possible alternatives.
3. Identify the possible outcomes.
4. List the payoff/profit or each combination of alternatives and outcomes.
5. Select one of the mathematical decision theory models.
6. Apply the model and make your decision.

Common decision making variables include cost and time. Most reports I’ve done in class ask for the cheapest alternative and the fastest alternative. Obviously that often times the best alternative is where both meet. There are, however, many other variables to consider. Without wanting to make this post longer than what an exposure should be, I’ll finish up with a closer look at step number 5, which is the most complex, perhaps, and where all the math rules come in.

Selecting one of the mathematical decision theory models

Selecting a theory model, depends largely on the environment in which you’re operating and the amount of risk and uncertainty involved. There are 3 environments.

Types of Decision Making Environments

1. Decision making under certainty – In this situation, there is certainty about the consequences of each alternative.
2. Decision making under uncertainty – Zero or little probability data available about each alternative. Methods include:
• Maximax
• Maximin
• Equally likely
• Minimax regret
3. Decision making under risk – There is significant probability data on the possible outcomes. Methods include:
• Expected monetary value (EMV)
• Expected value of perfect information (EVPI)
• Expected opportunity loss (EOL)