Mining Nuggets of Knowledge
How can you derive meaning from a chaotic world made of vast sets of data points? Well, you could turn to religion... Maybe even ideology. But if you're a marketer, you may want to try data mining first.
How can you derive meaning from a chaotic world made of vast sets of data points? Well, you could turn to religion... Maybe even ideology. But if you're a marketer, you may want to try data mining first.
Last week’s column focused on corporate data sources, including an aggregated shared data file as the basis for data mining and extracting customer data. The objective, of course, is to have enough disparate data points to allow the emergence of patterns that might not be obvious to the naked eye or to any of the independent points of customer contact within the organization.
The response from readers, though conceptually in favor of shared data files, was decidedly pessimistic about the likelihood of that level of sharing happening in their companies; technology enables efforts that are then thwarted by the human instinct to hoard and protect resources, including information.
Instead, the readers who wrote in asked that I start with an easier goal — garnering learnings from the data already accessible, particularly that information we all have in log files and Web-activity databases.
Based on the admittedly nonrandom samples of those who sent me email, I am shifting the direction of this column for the coming weeks to addressing issues around data mining, data analytics, and online analytical processing.
Data Mining
What is data mining? Simply put, it is the ability to derive meaning from large sets of independent data points. To marketers, data mining refers to any process that allows us to capture and utilize information more effectively, particularly with regard to customer interest, behavior, and profitability.
It is no longer enough to know how much we sold. We need to know which customers contribute what share of total revenue, which are most important to building a profitable business, and which market segments represent our current base, future potential, and greatest vulnerabilities.
The learning potential from smart data mining applications is endless; the challenge lies in asking the right questions and setting up your data warehouse to allow for endless flexibility as market conditions and customer behaviors change.
A number of reputable tools and solutions are on the market, with more appearing weekly as companies recognize the strategic importance of customer-centric data mining. And because this stuff is never simple, a number of different positions are being posed about how best to tackle these issues — that is, no two approaches are quite the same. Consider, too, that each of us has different business challenges, varying levels of customer interaction and institutional knowledge, and diverse legacy CRM systems and online data capture setups — and you can see why this whole arena has so many so puzzled.
A Business Necessity
But puzzled or not, the need is real, and the use of analytics is growing fast. The Aberdeen Group tells me that in 2000, the Web analytics market would reach $425 million, a 200 percent increase from the 1999 base of $141 million. Various pundits and soothsayers predict growth of anywhere from 10 to 100 times in the next five years.
Investing in systems to better understand customers (as well as employees, vendors, supply chains, etc.) is a modern business necessity. Ready or not, here it comes.
So we may as well get ready.
Next week, I’ll jump into the fray by looking at the big subcategories of solution providers and what those types of offerings can do for your business.