Why does data mining make sense for a company like mine (no pun intended)?
Many projects done with DM so far have been undertaken by companies that routinely collect high volumes of data as part of their ongoing business processes. These companies have found that it is not enough to simply collect and store the data, but there is also a competitive need to somehow try to make sense of it from a business standpoint. Some industries that fit this profile are retail (point-of-sale data), banking (transaction data), and telecommunications (call, billing and service transactional data). Typical applications of this technology to a financial services firm, for example, might be to create a credit scoring model or predict accounts that are likely to writeoff. Examples of business functions that would benefit from DM for a telecommunications company would be customer retention, fraud control and phone or computer network administration.Increasingly consultants and vendors are constructing "one size fits all" vertical applications for select industry segments, which they then, in turn, market to all competitors within that segment. A potential risk is for companies to explore DM only to the extent of simply buying these vertical "off the shelf" solutions that their competitors can also easily buy. However, applications of this technology are only limited by the imagination and need not be limited to only the above examples. In fact, the companies that will be positioned best to use this technology to gain sustainable strategic long term competitive advantage will be those that learn to "get comfortable" in turning these types of tools loose on all varieties of company data.
The true promise of DM is that it can help a company to continually learn from it's own data. It can utilize the data that has already been collected (usually at much cost) to learn about it's market, customers, environment, and potentially even it's own core competencies.
Why Should My Company Look Into Data Mining And How Should It Fit Into Our Overall Strategy?
Many Fortune 100 firms, as well as various consultant and vendor firms, are attempting to learn how to use this technology. The implication here is that those companies that don't learn how to use this new type of tool effectively within their organizations, may soon be left behind by those who do. Also, those that depend on outsourcing this "business intelligence" competency to consultants may find that they might be sharing their "competitive advantage" with their competitors.
What's The Best Way to Proceed with Looking into a New Technology like Data Mining?
The best way to start is to select a tool to work with on a "proof of concept basis" to see what types of advantage that DM can provide to your company. Taking the approach of learning to use these tools and applying them to the business should be taken in "small steps" and using a "pilot project" / "proof of concept" approach.
What is the Forecast for this Technology to Become Mainstream?
This technology is already becoming mainstream at a rapid pace (see the Success Stories below). As vendors continue to build vertical applications that incorporate data mining and make the technology easier for users to master, the technology will become even more prevalent.Success Stories of Companies That Are Using Data Mining Today:Does this type of tool sound "too good to be true"? It can be risky to be on the bleeding edge of technology, but it can also be costly to ignore potentially powerful tools that your company's competitors are looking into seriously.
Dutch Railway Company joins issue with train delaysBBC - Predicting Audience Share
Reuters - Data Verification for Foreign Exchange Prices
To send me feedback, click this: andrew.hall@anderson.ucla.edu
Back to main Data Mining index page:
Visual Stimuli courtesy:
Last Updated by AJx on 04/05/98 07:46:02 PM