Categorized | Data Mining & Analysis

Information Technology 6.8. Data Mining & Analysis

As your business collects information of all sorts, that information can be used help you make business decisions. Data mining and data analysis are the tools businesses use to turn information in databases into business decisions and business actions.

Despite its complexity and impenetrable codes and math equations, data mining and analysis is really very simple from your perspective. All data mining and analysis involves is asking a question of your data, questions like:

  • Who are my best customers?
  • I’m thinking of giving free shipping to encourage customers to order more―will it work?
  • I have a new product and want to send out an email to let people know about it―who are the best people to send it to?
  • I’m considering buying ads on a search engine and they will allow me to have my ads appear at a particular time of day. When do customers come to my site to buy so I can place my ad at just the right time?
  • What products are the most profitable?
  • Should I have a Christmas special?

It’s not like the database is going to come out and answer these questions for you. A database is only a recording machine. But, fortunately, there are a number of tools that allow you to mix up, match, combine, slice, and dice the data in your database. Those processes of mixing up, matching, and slicing/dicing data come under three major categories:

  • Database querying―all databases have a query language (usually SQL, for “Structured Query Language”) that allows you to combine data in a very simple way, such as, “Show me all female customers from Minnesota who have purchased something in the last week and who took advantage of our free shipping offer” or “How many people bought Widget A in September and how much did that change by November?” While writing these questions in the actual query language takes a madman and a genius, that’s basically what database queries are.
  • Data mining―data mining goes beyond querying data and uses sophisticated mathematics to group data, find trends, make predictions, and discover patterns. Data mining answers a question like, “When sending out direct mail to prospective customers, what type of person is most likely to respond?” Or a question like, “I have a prospective client who is in the manufacturing industry―how much profit am I likely to get from this client?”
  • Data analysis―data analysis takes what you learn from database querying and mining and presents the data in an actionable sort. A data analysis question might be, “Who are my best customers?” A data analyst would then sift through your customer data by querying and mining your database and present a quartile analysis that divided your customers into four groups from your most profitable customers (“platinum” customers) to customers you’re losing money on (“lead” customers) and all the relevant characteristics of each. Data analysis might show you that your best customers are female and over fifty while your “lead” customers have more than three children in the home.

In essence, you do database querying, data mining, and data analysis when you are attempting to turn information into something useful for your business, into a decision.

So you need two things. You need data, and enough of it. The more records you have, the better the data can answer your questions. The more detail there is in your records, the better the data can answer your questions. (You can see why, in section 6.6, I stressed that developing a database should start with the kinds of questions you want to ask of your data―if you’ve been using a database for a couple years that doesn’t tell you your customer’s gender and one day you want to find out if males or females make more profitable customers, you’re out of luck―you have a huge pile of data that can’t answer one of your basic questions.)

Secondly, you need questions. What do you want to know from your database? What kind of decisions do you need to make? These questions can be operational (who makes a good project manager?), financial, accounting, or marketing questions. As long as your data is sufficiently deep and broad enough to answer the question, your question will get a sufficient answer to guide your business decision.

For business that have a high volume of database transactions, data analysis and data mining are accomplished through the use of data warehouses designed explicitly for querying and analysis of data. A data warehouse is simply just another database; however, rather than data being added to it in a series of transactions, it retrieves specific information from a business’ databases every few hours or everyday. It’s designed only to house the information important for making business decisions. So, if your venture involves one or more high transaction volume databases, you need to build a data warehouse for any data analysis you wish to do. Many of the data mining services listed below specialize in building data warehouses to meet specific business needs.

6.8.1. Statoo Consulting

Based in Switzerland, Statoo Consulting is a software-vendor independent Swiss consulting firm specialized in statistical consulting, data analysis, data mining, and analytical customer relationship management.

6.8.2. Drayton, Drayton and Lamar, Inc.

Drayton, Drayton and Lamar, Inc. offers the following services:

  • Database and data warehouse development and support
  • Data extraction, transformation, and loading (ETL)
  • Data analysis
  • Special reports (e.g., congressional inquiries, Freedom of Information Act requests, and press inquiries)
  • Technical assistance
  • Technical documentation

6.8.3. JMP

JMP 7 is the most affordable and user-friendly data analysis, statistical analysis, and desktop data mining software available. It is a full statistics package that comes with very powerful data mining and analytics tools. It is, in fact, a scaled down version of the most powerful statistical and data analysis software on the market, SAS. However, unlike SAS, it does not work directly with your database; you have to import database tables into the application and save them as separate files―in essence, creating your own data warehouse on the fly. Available for Windows, Linux, and Macintosh, a single-user personal JMP 7 license, including one year of free technical support, costs $1,500 (I said it was the most affordable, not cheap). If data analytics is a very serious part of your business model, JMP can meet almost all your needs.

But don’t be fooled. You’re never going to pay $1,500 for an application that you can use right out of the box. The software demands that you know what you’re doing and, if you don’t, you should be prepared to burn the midnight oil for a few weeks and master the program and its underlying principles. If you don’t know what you’re doing, you’ll never be able to do the cool things it lets you do, like multiple regressions, clustering, and neural networks. For those of you unacquainted with statistics, the JMP learning curve is mighty steep, but once you get there, the power at your hands is unbelievable.

6.8.4. Analyse-it

Even if you don’t have a database, you can still do data analysis on any spreadsheets you might have (or you can export your database to a spreadsheet and analyze the data from there). JMP is the most powerful affordable tool for analyzing spreadsheets or databases, but you can also use Microsoft Excel at considerable savings, provided you jimmy the application a bit. Excel comes with some very basic statistical tools; you can up the ante by purchasing add-in software that gives you a pretty full range of statistical tools. While there are a dozen or so Excel statistics add-ins, the best and most popular of the bunch is Analyse-it. Like JMP, however, it’s relatively useless unless you understand statistics and know how to run them on data. The learning curve is steep, but once mastered, you’ll unlock the gold hidden in your spreadsheets.

Be Sociable, Share!

Leave a Reply

Shoestring Book Reviews

Shoestring Venture Reviews
Richard Hooker on Jim Blasingame

Shoestring Fans and Followers



Business Book: How to Start a Business

Shoestring Book

Shoestring Venture in iTunes Store

Shoestring Venture - Steve Monas & Richard Hooker

Shoestring Kindle Version # 1 for e-Commerce, # 1 for Small Business, # 1 for Startup 99 cents

Business Book – Shoestring Venture: The Startup Bible

Shoestring Book Reviews

Shoestring Venture Reviews

Invesp landing page optimization
Powered By Invesp
Wikio - Top Blogs - Business