Improve Workforce Retention by Using Data
By Bob Seemer, Chief Operating Officer, Electronic Training Solutions, Inc.
Employee retention remains one of the hottest topics, but strategies to improve retention often fall short. With so many organizations coping with staff turnover and vacant positions, why hasn’t this problem been solved yet? Could it be that managers are not getting to the root causes of turnover in their organization? Are they taking aspirin for their toothaches instead of going to the dentist to get their tooth fixed? Instead of using data, are they relying more on simple remedies or well-established industry-wide anecdotes which may have worked in the past, but no longer apply to a younger workforce or different working environments? Let’s review two case studies and try to shed some light on identifying root causes and finding the right, organization specific solutions.
Teams in two different, but similar sized, hospitals embarked on projects using data driven, Lean Six Sigma tools with the goal of increasing employee retention. Each organization had tried identical strategies to understand and address the issue in the past, but were still dealing with 24% and 27% annual turnover levels for workforces of more than 6,000 employees. Following are some of the key elements and findings from these two projects, starting with clearly defining the problem they were trying to solve.
Defining the problem
Should they focus on staffing levels, filling vacancies quickly, or turnover? Although these are related, my suggestion is always to first “stop the bleeding”. For these teams, that meant focusing on employee retention and, critically, defining the term “retention”. Data driven approaches require agreement on definitions of measures. In these projects, the organizations decided that a successful hire was one that was retained for at least 12 months after the hire date. With that definition in mind, each team then collected a year of data and developed visuals depicting their current performance levels (team #1 – 24% and team #2 – 27%), organization targets of 14% and 17%, and the resulting gaps of 10 percentage points each. Each team then developed a SMART objective statement. For example, “Reduce the annual employee turnover rate from 24% to 14% by December 31, 2019, to impact the organization’s strategic goal of a “Highly engaged workforce”.
Calculating the cost of status quo
Along with the performance measure, target, and resulting gap, each team had to quantify the financial impact, or the Costs of Poor Quality (COPQ), the gap was having on the organization. Roughly speaking, what was the financial burden for each hospital because of the need to hire more employees each year than the targeted level? For example, if each had an authorized staffing level of 6,000 employees, and the gap was 10 percentage points, then 10% X 6,000 = 600 more hires than the target. If each new hire costs $20,000 to recruit, onboard, train, and support administratively, then 600 X $20,000 = $12,000,000 annually. With the gap of 10 percentage points and a COPQ of $12 million annually, a compelling business case was able to be presented to each leadership team.
Analyzing the right data
This can be the most challenging step for organizations as project teams and leaders are frequently eager to skip from defining the problem and its costs straight to implementing solutions. In our case study examples, each organization had relied on employee climate surveys, exit interviews, and anecdotal stories learned from industry conferences and publications to address retention. The problem with these types of data is that they tend to be opinions, unreliable, or fables. Surveys, or the analysis of the resulting responses, are rarely done with the academic rigor needed to rise above opinion. Not every departing employee has an exit interview and many interviewed employees won’t tell the truth for fear of reprisal. Stories and even “best” practices are frequently conveyed in a data-free manner, making it difficult to determine the applicability to individual situations. So, what kind of data are the right data?
The right data are specific to the processes of your organization. In these examples, the teams had to define and understand the process producing the poor performance. In this case, teams documented the “journey of a new hire” which started with “Vacant Position Identified” and concluded with “Vacant Position Was Filled for 12 Months”. The flow of steps for “the journey” was captured showing decision points, participating departments, and potential trouble spots. This helped inform the best approach for collecting reliable, organization-specific data, which began with determining the tenure of employees who had departed in the past year. Since a successful hire had been defined as one with at least 12 months’ service, the teams focused on those who had failed to meet that standard.
For team #1, the analysis revealed that 4% left within 30 days of hire, 22% within 90 days, and 53% within 12 months. This directed their attention to two cohorts – the 22% who left within 90 days of hire, and the 31% who left between 91 days and 12 months of their hire dates. Next, they analyzed the demographic data for the two cohorts and found that RNs most frequently exited. This knowledge enabled them to develop targeted SMART Problem Statements that, if successfully solved, would have an impact on the overall objective to improve retention. For example, “Reduce the number of RNs leaving between 91 days and 12 months of the hire date by 50% by December 31, 2019, thus reducing the overall employee turnover rate from 24% to 19%”.
The second team followed a similar process, but, critically, did not have the same result. For team #2, the data revealed that RNs were not the significant leavers, but rather Millennials. This was a surprising result, but indicative of the power of data and the importance of investing the time in analysis specific to your organization.
Verifying the root cause
Once the right problems are identified and well-defined, verifying the root cause is easy. Cause and effect analysis, 5 Whys analysis, and frequency analysis indicated that each problem shared similar root causes, but also had unique ones. For team #1, employee selection criteria and interviewing procedures were unique root causes to the 0 – 90-day cohort, and workload allocation and poor mentoring procedures were unique to the 91-day – 12-month cohort. Understanding these differences allowed the teams to select the solutions most likely to be successful in solving the problem.
Implementing robust solutions and documenting results
Now we get to the step that people are most tempted to skip ahead to – implementing solutions. In addition to skipping the analysis and root cause verification, there are a multitude of reasons improvement initiatives fail in the implementation of proposed solutions, regardless of the potential benefits. The risk of failure can be mitigated by the application of project management techniques including Work Breakdown Structure, Risk Analysis, Action Planning, pilot testing and calculating the expected Return on Investment. The teams applied these tools, so implementation could be properly managed and reliable estimates could be made regarding the expected benefits.
Countermeasures (i.e., solutions) for team #1 were implemented about 90 days after project initiation, and preliminary results were documented after 90 more days, just as team #2 got started. By the 180th day, new hires leaving between 0 – 90 days of hire dropped by 65%, well below the target. A small decline in the overall turnover rate from 24% to 22% was also documented. After another 180 days, both team #1 cohorts were showing significant improvements, exceeding the targets set in their respective Problem Statements. By the close of 2019, team #1 had achieved its target of 14% overall turnover and therefore, eliminated the $12,000,000 Cost of Poor Quality.
Team #2, which started its initiative 6 months later, also showed significant improvement, but was impacted by the start of the pandemic in early 2020 and was unable to isolate its improvements from external conditions affecting all aspects of its organization’s operations.
Maintaining the gains
Sustainability is challenging for most process improvement initiatives. It is key that process changes are supported by revised standard operating procedures, documented workflows, training and mentoring so that solutions become “standardized”. Otherwise, performance gains will gradually decline to previous levels.
Summary and Lessons Learned
It is tempting to follow the lead and rely on industry publications, experience, and anecdotal data from interviews, surveys, and experts when searching for solutions. Although all organizations may be composed of systems, processes, people, and equipment, how these elements work together is always different, thereby producing different outcomes for each organization. Just because two organizations may share the same poor performance level doesn’t mean they share the same problems or root causes. The best approach is to apply a systematic approach for process improvement, collect the appropriate quantitative data, and utilize the appropriate analytical techniques to develop the right solutions for your organization.
Bob Seemer, Chief Operating Officer, Electronic Training Solutions, Inc.
Bob is a management consultant and trainer for organizations in all sectors. He specializes in Baldrige-based organizational assessments, strategic planning, and process improvement utilizing Lean Six Sigma, Project Management, and Innovation methodologies. He has more than 50 years’ experience and lives in the Atlanta area.