As companies are juggling more data, balls are flying all over the organization. Statistical analysis is no longer a mysterious art performed by the advanced gurus. Its powerful benefits are increasingly apparent to the average business user and adoption is growing steadily. Analytical software provider SPSS announced the release of its Statistics 17.0 today -- and renamed the product SPSS Statistics to differentiate it from the company itself, which distributes a wide variety of products. The new version aims at addressing four industry needs, according to information provided by SPSS:
According to Marcus Hearne, manager of product marketing at SPSS, the improvements made in version 17 were incorporated equally from two sources: customer demand and the strategic concerns of SPSS itself. (One question the company asked itself, he says, was "Where will the technology be in 10 years?") The emphasis with this release, he says, is on providing novice and experienced users with a single tool that fits the needs of the entire enterprise. Before, basic users were trapped in spreadsheets such as Microsoft Excel while advanced users toiling away in programming languages such as C++, making it difficult for either side to communicate effectively with the other.
- better tools for research and reporting to address new business problems, new data types, new output formats;
- ease of use for beginners and greater functionality for advanced users;
- better visualization tools; and
- integration with existing SPSS products (e.g., data-mining tool Clementine) and other vendor tools (e.g., data stored in Oracle).
"What we've seen, as the availability of data becomes greater, [is that] more and more people are saying, ‘I want a better way to digest this data,’ " says Rebecca Wettemann, vice president at Nucleus Research. SPSS itself was seeing more and more generalist users -- those who had outgrown Excel, along with undergraduate- and graduate-level students and business users. Given a wider demographic, SPSS also changed some of the names of its modules to make it easier for users to understand their purpose. "SPSS Trends," for example, might make sense to a core group of users, but the company has changed the module name to "SPSS Forecasting" to make it more recognizable to a broader audience.
Hearne highlights exciting new research tools such as "multiple imputations for missing data," where users can "safely and securely" fill in gaps in their existing data. Given the astronomical amounts of data available to a company, the amount of missing data also increases. "It’s almost impossible to get all of the data, all the time," Hearne says, "Something’s always missing or wrong." Another tool is "nearest-neighbor analysis," which allows companies to locate information that might otherwise have been overlooked: A given data set -- about competitors, prospects, customers, etc. -- can produce a list of similar items that might be deserving of some attention, given their resemblance to the ones already identified.
In terms of reporting, SPSS says it has improved the process by allowing users to format their data within the application and then export it to any number of external programs (e.g., Microsoft Word or PowerPoint). This saves users the trouble of having to go back into Word or PowerPoint and reformat, a task that can be especially cumbersome when it comes to large projects. SPSS's improved visualization tool, Wettemann notes, "is a great step toward making everyone think about data in a structured way for decision-making."
To increase the use and adoption by beginners, Hearne points to ongoing efforts to improve the user interface. Moreover, functions like recency, frequency, and monetary analysis (RFM) have been added -- an application that’s very useful to a marketing staffer, but one that might sound foreign to a statistical engineer.
For advanced users, SPSS is creating a platform that’s "open to anything and anyone," Hearne says. With the availability of what the company calls its "programmability extension," users can call upon external programming languages, such as R and Python, and customize logarithms that fit their own needs. In addition, its Custom Dialogue Builder gives advanced users the freedom to create the syntax, and then essentially translate that syntax to match user skill sets and responsibilities, according to the company. The goal of each release, Hearne adds, is "doing what's right by the majority of users," which may mean that the company can’t create a solution for every need. "We’ve created so much customizability [to allow them to] create it themselves," he says. "We try to remove any roadblocks. We don’t want users to wait for us to catch up to them."
While awareness of the power of statistical analysis may be high, successful applications of the approach have yet to reach the same level of fervency. Most organizations, Wettemann says, are using some form of statistical analysis, but it mostly tends to be isolated within siloes. Moving forward, analysts and vendors will have to measure adoption rates based on:
No doubt the process is being made less painful by analytical software providers such as SPSS, which Wettemann points out is low cost and flexible. Nevertheless, she adds, "I think everyone in the market could make their solution even easier to use. We'll continue to see pressure in that direction."
- how many people are actually using statistical analysis; and
- how often they’re using it on a regular basis.