Big data, cows and cadastres
The company has rolled out Pentaho Business Analytics 4.5. That version features interactive visualization and data exploration capabilities that access all data sources. The system supports big data and offers licensees an extensible interface for software and cloud-centric companies to expand the base systems with third-party visualizations.
Between the giants like IBM and the upstarts like Pentaho, dozens of startups and smaller firms are eager to snag a ride on a big data bull. Software is becoming easier to use. The reason is that the Kolmogorov-type of math ability is in short supply. Software can help a person with little or no math ability make sense out of big data. I am not 100 percent on board with the idea, but user-friendly analytics for big data analysis is a booming field.
One vendor, Tableau Software, offers one of the most user-friendly big data analytics systems available. With a few mouse clicks, Tableau will output PowerPoint ready charts and graphs. The challenge is not making it easy to click and get a report.
Consider this example. A midsize company has a website that allows employees, customers and prospects to request information. The marketing manager responsible for that particular section of the organization's website is not receiving requests for information. Amidst the thousands of data points generated by the website's log file system, the marketing manager is looking at zeros. What do you do? The answer to this question is difficult. If the company has an IBM system, the process required to get data that answers the marketing manager's question may require getting a task in the queue, then a discussion with a business analyst, a wait of a day or a month depending on staffing, and then a report. If the report does not answer the question, the cycle must be repeated.
Using one of the user-friendly desktop or cloud services, the marketing manager can navigate via a browser to a setup screen. The manager points at a data set and clicks on a report, and the system delivers data in tabular or graphic form. The only problem is that the user-friendly system assumes that the marketing manager understands sample size, the strengths and weaknesses of specific statistical methods and the output itself. Eye-catching graphics is not the same as statistically valid data.
The problem in those two examples boils down to people. There is a shortage of staff with big data and analytics skills. The problem is not local; it is global. Data and the need to exploit it are rising faster than the talent pool required to use the sophisticated, increasingly user-friendly systems. Kolmogorov worked with a pen and paper. He could tap into today's powerful system because he had the mathematical expertise required to tame big data. Using a mouse is the trivial part of figuring out cow genetics.
The Russian tax authorities, the dairy industry and managers can benefit from big data and analytics. Those who misunderstand—including marketers who don't know a valid data set from a bovine RFID tag—could end up on the barbecue grill after the rodeo ride ends.