Statistics in the Workplace Use of Statistics

¶ … Statistics in the Workplace Use of Statistics at Cincom Systems The use of statistics in pervasive throughout Cincom Systems, from tracking customer behavior through research surveys to measuring software quality and continually striving to attain higher levels of application performance and stability. Statistics are specifically used for managing the ongoing investments in online marketing including Google AdWords analysis, where descriptive statistics are used for measuring the performance of specific programs and strategies.

There are also inferentially-based statistics used for measuring the overall level of software quality today and what levels would need to be attained to keep on track for having all applications available on the Web via the Software-as-a-Service (SaaS) model by 2015. Statistical Analysis and New Product Development Using ongoing customer surveys and market research programs, much data has been collected on what customers want in future product released, and how important each of these features are.

There is also much focus on defining which platforms to migrate specific software applications to as well, with Apple IPads being a favorite of the most forward-thinking customers. All of these insights were gained using descriptive statistics including frequency distributions, cross-tabulations or joint frequency distributions and uni-variate analysis. Every three months these statistics are shared with our senior management team, which uses them to define how the company will respond to rapidly emerging new markets.

The use of statistics as a means of gaining a competitive advantage in new markets is a highly effective strategy in congested and highly competitive markets (Javalgi, Gross, Joseph, Granot, 2011). There is also the use of descriptive statistics for the planning of updates and refreshes to existing products, and this has proven to be very effective as a product refresh strategy in the process. The listing of items, weighted and scored using descriptive statistics, provides user groups and the largest customers with insights into future development.

This has transformed the product launch process internally, providing much greater levels of information than had been available in the past as well. Where marketing, product management and product planning are concentrating on descriptive statistics, software quality assurance is concentrating on the use of inferential statistics to project how many users will encounter a specific software problem and need a revised application. This use of inferential statistics has also helped to significantly reduce the time to fix bugs or problems in the software code as well.

An example of this is when a quality audit showed that a specific code segment from an Enterprise Resource Planning (ERP) application was no longer compatible with integration code and techniques used in the past. The inferential analysis also showed that only ten customers would be affected by the change to ensure the ERP system platform could still work with other systems. Engineering, product management and services all coordinated on a rapid turn-around on the code and the problem was fixed in less than a month.

The accuracy and speed of the fix to the legacy ERP system code would have not been possible without the use of inferential statistics. The four levels of measurement encompassing nominal, ordinal, interval and ratio scales are also extensively in software quality management for defining the accuracy of quality measurements (Champ, 1995). Interval and ratio-scale data is often used for determining the magnitude of the software quality update or change necessary for an application to perform effectively.

Nominal and ordinal data is often used to determine relative levels of user adoption of a given feature or specific area of concern in a new system (Brandt, Henzen, Ramos, Bucchianico, 2007). All four levels of statistical measurement are also used intensively in customer surveys and the analysis of customer data as well, to delineate specific segments and their unique requirements or needs (Javalgi, Gross, Joseph, Granot, 2011).

One marketing director has used interval and ratio-level data to define the specific subsegments of a given market and define which product attributes or features were the most critical to each given audience for the product. Prioritizing specific features, how they relate to given segments, and also creating specific programs based on the relative weight or value of one attribute or another relies on interval-level data analysis.

The data collection methodologies necessary to attain this level of insight are significant and require research projects to be highly tailored to specific strategies and initiatives in the business (Javalgi, Gross, Joseph,.