Better BI

Chris Gerrard – Exploring the world of Better Business Intelligence

A Note on the History of BI

with one comment

Way back in the old days business intelligence meant data processing and reporting.

The devices for publishing business data were line printers and character-based terminals, usually of the IBM 3270 family.

Reports were pretty straightforward affairs: appropriately collected, filtered, sorted, and aggregated data was presented in tabular format, with no or very little formatting.

This worked well for addressing a very large variety of business information needs (and remains an excellent solution for these needs today, albeit with vast improvements in formatting).

Analytics were restricted to varying the elements of the reporting process. Record selection, varying the contents of a report by including or excluding records based on data filtering, was the most common, permitting the same “report” to be generated with different data populations. This made is fairly straightforward to compare data over time or across meaningful business dimensions e.g. divisions, regions, product lines, etc.

Changing the structural characteristics of a report – reordering columns or changing sorting or aggregation, makes it very difficult to compare like-to-like. Providing mechanisms for changing the structure of reports has been possible for as long as reporting tools have existed, but for a long time the practice of structurally altering reports was not widespread.

The advent of non-mainframe computing environments introduced all sorts of new reporting capabilities.

In particular, graphics-capable computing led to the advent of basic business graphics. Bar and line graphs permitted the visual presentation of data in ways that permitted the observation and interpretation of quantitative values of and relationships between similar data sets.

Over time the ability to produce a wide variety of graphical forms, with increasingly varied visual attributes, grew. And grew, and grew, and grew.

Pretty soon we were awash in data graphing in all its glory. And gory misapplication of visual effects.

For although there is a long history of study and practice in the visual presentation of quantitative data the practice of presenting data has languished in a world much like the early days of the World Wide Web when everyone who could sling together some HTML took advantage of all the visual features and created truly hideous visual monsters. Well, at least a fair number of them did.

Over the years there have been some excellent works published that provide good, solid, practical advice on the presentation of data in a business environment. Edward Tufte’s works are the bedrock upon which most of the modern best principles and practices rest. Stephen Few’s books are marvelous in their sheer practicality in providing good solid examples of how, and just as importantly how not, to present data.

Where once we were restricted to single-report production and publication we are now faced with an embarrassment of riches. We can dynamically construct reports of dizzying sophistication and subtlety with interactive capabilities that let us scour the full range and depth of the organization’s data. We can build dashboards that faithfully mimic the cockpit of a Formula 1 racecar, and all too often do (more on this in future installments).

We can create executive summary monitors that elegantly, completely, concisely, and cogently present the full range of relevant data necessary for a business decision maker to understand the full spectrum of the state of his business. Some would call these dashboards, but that word has been usurped and polluted by the marketeers, the Formula 1 cockpit jockeys, and has lost all of its dignity.

In this age of Big BI we are faced with an entrenched paradigm of monolithic vendor-driven enterprise-scale highly complex all-inclusive technology platforms that are thought to be (because they’re promoted to be) the One True Way for implementing Business Intelligence in the modern organization.

These behemoths are explicitly designed to consume and digest all of an enterprise’s critical business data into Data Marts and Warehouses and provide the full suite of analytical and publication facilities that satisfy each and every one of the needs of anyone who needs to understand the organization’s data.

The sad reality is that from the only perspective that really matters – delivering high quality actionable business intelligence (information) to business decision makers – the large majority of Big BI initiatives fail miserably.

This blog is my address to this state of BI, and explorations in how to do it better.

We, as Business Intelligence professionals, can do better.
We must do better.
We can do better.
We shall do better.

Chris Gerrard


Written by Chris Gerrard

February 4, 2009 at 12:45 am

Posted in Uncategorized

One Response

Subscribe to comments with RSS.

  1. […] Its passing was inevitable. Big, bloated, and overripe, it never lived up to its promise of being the path to data enlightenment. Although its nominal goals were good and proper, in practice it failed to deliver. Much has been written on Big BI’s problems, including here, here, and here. […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: