|
By Carel Badenhorst, Product Manager Business Intelligence, SAS South Africa
I have never much of one for myths and superstition, which is surprising considering that my strict Afrikaans upbringing was littered with these as well as a handful of old wives tales, a possible side effect (like the fourteen child families) of long, dark farm evenings without SuperSport on DSTV. My personal favourite of these is the one that says "If you lie down and eat, you will grow horns out of your head".
From a young age, some things I was told I took for granted as being honest and true without question - like my dad telling me he would pull off my ears if I did this or that again. For the most part however I have been a tester of theories all my life, including most of the superstitions I have been fed as "being the truth" as well. Numerous ladders have been walked under, several mirrors have bitten the dust, and I have lain on the couch on a Sunday afternoon eating biltong without experiencing horrible bad luck or growing a single horn on my head.
Even though in my daily business life 1 rarely have to deal with black cats crossing my path or throwing a pinch of salt over my shoulder, I am still faced with several rather interesting business intelligence myths which I feel deserve to be thoroughly 'busted'.
MYTH#1
"With data warehouse implementations, it will take you a millennium or more to achieve any sort of Return on Investment and business value"
The Truth
The only difference between a tedious and potentially low business value data warehouse implementation and a to-the-point, quick potentially high business value data warehouse implementation is simply planning. Like planning the building of a new house, data warehouse implementation planning includes various disparate activities answering the questions like: what should I do, why should I do it, when should I do it and how should it be done? All of these questions will be answered quicker if a detailed blueprint exists to guide the activities. The SAS Rapid Warehousing Methodology is such a preexisting blueprint. It has been tried and tested in numerous implementations; it works and achieves quick, iterative, high ROI data warehouses.
MYTH#2
"Data Warehousing can and will never be able to deliver business benefit in operational real-time"
The Truth
With industry leading technologies the concepts of Customer Data Integration (GDI) and Master Data Management (MDM) is a reality in the SAS world, helping you achieve a fully integrated, single, trusted view of your customer data or any other operational data you need to do Intelligence on outside of the normal data warehousing arena in operational, real-time.
Not only can you get this view of your customer (GDI) and other operational (MDM) data in real time, it also allows you to do operational synchronisation. Once you are happy with the single, trustworthy view of your data you are getting in real time, you are able to write back to any or all of your operational systems to ensure that all your operational sources contain the same trustworthy copy of the data.
MYTH#3
"Achieving Data Quality is a great long term goal for Business Intelligence, but not imperative to have from the start to achieve Business Intelligence"
The Truth
Achieving Data Quality is an imperative first step for Business Intelligence success. The old adage of GIGO (Garbage In, Garbage Out) means that Business Intelligence systems will unquestioningly process the most nonsensical of input data and produce nonsensical output, which inevitably will lead to failures in humar decision-making based on this faulty, incomplete and imprecise data. At SAS we recognise this potential shortcoming in the Business Intelligence process, and have developed an integrated Data Quality Solution as integral part of our Business Intelligence platform.
MYTH#4
"Online Analytical Processing (OLAP) is Analytics and will help you drive your future business strategies "
The Truth
Online Analytical Processing (OLAP) is not analysis. It is primarily a way of storing data in a clever way to allow quick, easy access to huge amounts of data in a way users understand. If I sell sea shells on the sea shore in December, and I want to investigate how many sea shells I sold for that time, I would have three dimensions involved in that query, namely sea shells (product), on the sea shore (geography), and December (time). Traditional tabular (row and column) data layout has space for only two dimensions at a time for viewing queries, but OLAP will allow me to add the third dimension to my tabular query display and still receive a single, understandable, view of the query. Thus I achieve the ultimate of viewing my data.
Now imagine a business where we have thirty, or a hundred business dimensions. To achieve this single understandable view over all these business dimensions, I'd use OLAP storage. There is a second part to this statement that OLAP is a storage mechanism, not analysis, for it will only ever enable me to answer questions about my historic data past, or my data present, thus at best OLAP will provide me with Business Information, NOT Business Intelligence. I will never be able to start asking intelligent questions about my business future.
True Analytics like regression, ANNOVA, correlations, neural networks, decision trees, segmentation, credit scoring, forecasting and many others allow me to truly have an in-depth understanding of my business past, my business present and also what is likely to happen in my business future and what I can do to start preparing for that. This true analytical power puts the Intelligence in Business Intelligence. Couple this analytical functionality with good data quality, and you are a good way down the road to achieving true Business Intelligence. "If you can do pretty pictures you are doing Business Intelligence"
MYTH#5
"If you can do pretty pictures you are doing Business Intelligence"
The Truth
This is the least time consuming, yet from a user perspective, the most important part of any Business Intelligence system. This is the query and reporting (pretty pictures) interfaces our users will see and place so much focus on and will understand Business Intelligence to be. Yet those in the know realise that the hard work pertaining to achieving Business Intelligence is not the query and reporting, but reside within the realm of data access, data quality, data integration, data storage and data analytics.
The query and reporting interface is just a means to an end (the end being true Business Intelligence), not an end in itself. Query and Reporting has three main objectives in mind:
- It is a nice way of presenting true intelligence (refer to OLAP myth above) in a quickly understandable tabular or graphical way to enable decision makers to align their decisions to their business strategies;
- It should put the query and reporting power in the hands of the end-user, simplify their interaction with their data in a way they understand and can take action on, and as we all know empowered users are happy users which assures usage and buy in for Business Intelligence; and
- If done correctly, it will shield our users from the true complexities surrounding the process of achieving Business Intelligence.
Irrespective of whether the intelligent decision making in your business is being hampered by Business Intelligence myths and superstition or not, there is only one appropriate question to ask: "Does your intelligent decision-making rely on a proven, trustworthy, pre-integrated, enterprise wide Business Intelligence Platform enabling you to efficiently commit all activities associated with true end-to-end Business Intelligence?"
SAS Enterprise Intelligence Platform (EIP), helping you effectively deal with any Business Intelligence myth and superstition keeping you awake at night.
Contact us on: Johannesburg +27 11 713 3400 Cape Town +27 21 912 2420
|
Features 
Press Releases
