This is the second in a series of articles reviewing innovative ways that companies are leveraging Internet technology to offer new and better forms of staffing systems. Please note that neither of us are compensated in any way for our decision to write about a specific company or product. In this article we discuss how Unicru uses the data collection capabilities of the Internet to actively manage and improve staffing processes. Unicru’s system continuously collects and analyzes data at different stages in the employment life cycle, starting with the initial job application and ending with a post-exit review of the performance of former employees. Unicru uses this “streaming data” for two purposes:
- The data allows Unicru’s staffing assessment tools to learn from experience. As a result, the longer Unicru’s system is in place the more effective it becomes for identifying high potential candidates.
- Unicru provides clients with ongoing reports containing information to more effectively manage their staffing processes.
Unicru’s approach represents an initial implementation of what we believe will become a fairly standard practice, as online staffing systems evolve to more fully leverage the capabilities of the Web. As such, looking at Unicru’s approach to data management provides a good glimpse into the future of online staffing assessment. Unicru System Overview Founded in 1987, Unicru began by selling staffing systems primarily designed to automate the hiring process for high volume, hourly positions. Unicru was one of the early pioneers in the use of web-based staffing, and has developed a strong client base representing a variety of large national and regional retail organizations. Although Unicru offers a variety of staffing systems, they are probably best known for their ability to support the staffing needs of geographically distributed businesses, such as retailers, through placing web-enabled kiosks and other data-capturing devices in a client’s stores or offices. These kiosks allow applicants to electronically apply for jobs onsite, greatly reducing the time hiring managers spend screening candidates and enrolling newly hired employees into the company’s HRIS or payroll system. The initial benefit of Unicru’s system is increased applicant flow and decreased time-to-hire. For example, Blockbuster’s use of Unicru’s system resulted in a reduction in time to hire from several weeks to 48 hours. While these gains in efficiency are impressive, there are a variety of applicant tracking and candidate management vendors who can also boast similar sorts of accomplishments. Consequently, this does not reflect what we feel is the truly innovative aspect of Unicru’s system. What makes Unicru’s system unique is not how they automate the staffing process, but how their system leverages data collected from applicants and employees. The following represent the key steps in Unicru’s system:
- Applicants enter a variety of data into the system as part of the application process. This data includes general information, such as address and basic qualifications, as well as information about employment history, interests, and work styles that can be used to predict job performance and tenure.
- Applicant information is processed and fed back to hiring managers within five minutes after it has been uploaded onto the Web. The information is presented so that it can be used to effectively support hiring decisions.
- As applicants are hired, their records are updated online by hiring managers and the information is used to populate hiring documentation and employee records in the database.
- As employees terminate, the database is updated again via payroll data feeds and online exit surveys conducted by managers and departing employees.
The result of these transactions is the creation of a “closed-loop employee record.” This record contains a summary of the complete employment record of each person hired through the system. On an average day, Unicru will process between 17,000 and 21,000 employment applications. The Unicru data warehouse currently contains 14.8 million employment applications and 1.7 million closed-loop employee histories. Unicru uses this data to provide clients with reports giving insight into hiring practices, employee retention, productivity, headcount, and other staffing issues that directly impact company performance. These reports go well beyond the standard EEO and system usage reports offered by most staffing assessment vendors. The data also allows Unicru to make ongoing improvements to the design and scoring of their staffing assessment tools. Unicru has developed a number of unique analytic capabilities to process this data, and has patents pending on several complex artificial intelligence and data mining techniques. Does It Work? Given Unicru’s data intensive nature, it comes as no surprise that they have done several in-depth analytical studies into the effectiveness of their staffing systems. A recent study of data collected across several companies showed that Unicru’s systems were associated with a 42% increase in median employee tenure, a 20% decrease in involuntary terminations, and a 10% increase in employee rehire eligibility. After implementing Unicru’s system, one client saw a 25% decrease in its hiring activity. This translated into 55,000 fewer hires per year. Unicru is so confident in the effectiveness of their system that it is willing to “stake its fees” on its performance. Some of their contracts link the fees they receive directly to tenure or other business improvements achieved by a customer. Why Does It Work? In our opinion, there are two basic reasons why Unicru’s system works. First, they start with solid design. The content found in Unicru’s system is a result of years of experience and research into staffing methods. While Unicru’s current system uses technology that did not even exist prior to 2001, the basic principles underlying the design of their system and the assessment tools it contains are rooted in years of psychological research. In essence, they do not attempt to substitute “technological flash” for good science. Second, they “hedge their bets” by collecting and making use of streaming data. This data allows Unicru to continuously refine and improve the effectiveness of their system after it has been deployed. This represents a radical departure from more traditional staffing assessment design, where selection tools are developed during an initial study and then left in place “as is” for years and years without any major changes or revisions. The main challenge to using Unicru’s system is getting hiring managers to faithfully and accurately enter data on employee performance, thereby “closing the loop.” To help ensure this happens, Unicru has established training programs and a complete customer service outreach program to promote hiring manager satisfaction and use of the system. However, getting hiring managers to take on such additional actions at the request of human resources is not likely to be easy. Unicru’s system should probably only be implemented when there is strong leadership support to drive the internal behavioral changes needed to ensure its effectiveness. However, given the clear financial benefits that Unicru can demonstrate, getting such leadership support may be easier than one might expect. Unicru’s system is also designed specifically for positions that generate large amounts of applicant and employee data. Most of their clients hire thousands of employees per year. In addition, some of the statistical methods underlying the scoring of their assessments require relatively large sample sizes to be fully effective. As such, the system may not make sense for applications involving smaller staffing volumes. Conclusion In addition to Unicru, there are a few other assessment companies that actively recommend the use of ongoing “streaming” data to revise and improve the effectiveness of their staffing systems (for example, ePredix, Career Harmony, and FurstPerson). However, Unicru is the only company we know of that has made the collection and use of streaming data an integral a part of their solution. Although Unicru’s use of streaming data is still quite rare in the market, we feel it represents an approach that will eventually become commonplace as companies stop viewing online staffing assessment tools as static “tests” and begin seeing them as dynamic measures for guiding real time management of human capital.