In early 2022, the new leadership at Poll Everywhere set a course for the company to try a four-day work week (4DWW) experiment that we called “Summer Fridays.” We set the expectation for our entire company – approximately 60 employees – to work 32 hours a week across only four days. Most employees took off Fridays, however, some customer-facing teams staggered the days off to provide service for our customers. After several company wide surveys, statistical analyses, and considerable thought and discussion we have arrived at some conclusions about how a four-day work week suited our business, culture, and employees.

I have already discussed our conclusions and why we conducted the trial. Here, I will elaborate on how we see the problem and how we evaluated the costs and benefits of a four-day work week for our business.

Methodology

We reviewed the three employee surveys, had discussions with many of our managers 1:1, talked through the experiences from Ops and Recruiting, reviewed our own experiences, created a statistical breakdown of the questions, and pulled the typical quantitative data relevant to company KPIs, project deadlines, and performance for the period in question. We then engaged in the Bulletproof Problem Solving (McKinsey) methodology for breaking down a problem and iterating on the problem definition, structure of the variables, analyses required, and conclusions. We focused on our goals of employee retention, mental health, productivity, accountability, and retention for this analysis.

MECE process for the goals

The breakdown into a more mathematical and “Mutually Exclusive” and “Collectively Exhaustive” or MECE set of concerns sorts a few of our goals into the same buckets. We considered the 4DWW claims of improved Employee Engagementas something ultimately captured by examining productivity.

We see the 4DWW claims of improved creative work quality and output as very likely to be true based on our data, however, this is also ultimately captured by examining productivity.

We observed a minimal change in recruiting and made similarly small in-period efforts to attain PR coverage for this experiment. We spent so much on entirely unrelated recruiting investments in this period that we recognize this factor as potentially confounding. Additionally, recruiting can be seen as a decreasing cost in the case of improving Employee Retention. Thus, recruiting is a minor variable in the analysis.

Further, we see that our examination of Mental Health accrues benefit to Poll Everywhere only in the sense that it makes Poll Everywhere a better place to work and thus improves Employee Retention. We do not wish to minimize the benefits to individuals in this case, but the analysis proceeds to view the issue through Poll Everywhere’s interests. This is how we evaluate any benefits that Poll Everywhere may offer employees.

It bears mentioning here that a 4DWW experiment offers very few, if any, substantive levers that could move revenue aside from potential productivity improvements that we did not observe. We must then examine the impacts on costs at the highest level, balancing the costs of lost productivity with the benefits of improved employee retention.

Productivity

We arrived at an estimate of 10-15% lost productivity. This is still less than the 20% loss of total work hours and represents some progress. This estimate is driven by data like missed project deadlines and anecdotes from multiple managers about reduced capacity for collaboration and execution due to the lost time. The group we expected to most benefit from this experiment, in terms of creative work output leading to productivity, was negatively impacted and struggled with the lost time to collaborate in Engineering. The customer-facing teams struggled with the changes a bit less than anticipated.

Employee Retention

It is typical for HR departments to estimate the cost of a lost employee at 1-2 times the employee’s annual salary. This considers the opportunity cost of lost productivity, the cost to train, the cost to rehire, and some marginal costs for lost knowledge.

It may not factor in morale, but for our quantified purposes here, morale would ultimately only be represented in the retention of other employees or changes in productivity.

In this scenario, with reduced productivity, to expect to break even on an annual 4DWW, we would need for all of our employee turnover to decide to stay substantially longer with the company. Unfortunately, in our analysis, we do not believe that 4DWW would have impacted the choices of these individuals considered. The dominant concerns expressed were mostly personal preferences outside our control.

Other considerations

We also considered other risks that we found difficult to analyze or quantify but that deserves some consideration or mention.

Lesser relationships in Success

A loss of time may result in lower quality relationships.

Loss of high performing employees

There is a perception that American work culture views 4DWW as ‘more relaxed’ rather than a smarter way of working. This perception may cause the most career-growth-oriented to pursue work elsewhere. Growth is our stated ‘why.’ Primarily a concern for new hires in the future, but it may impact existing employees.

Only shared time off impacts happiness

There is a Swedish study on this subject with PTO and the overall consumption of antidepressants as a proxy for happiness. The study concludes that offering discretionary PTO days that are not broadly shared does not improve happiness for Swedes. Many of our own employees hinted at this outcome in that they did chores around the house on their extra day off rather than pursuing something else with loved ones.

Hedonic adaptation

The positive benefits will wear off quickly and carry the costs more than the benefits. This is a well-known experience common to human psychology of accruing any improvement in circumstance. The improvement becomes the expectation.

Conclusions

We arrived at the idea that we need improvements in Employee Retention to offset some costs in Lost Productivity to pursue an expanded 4DWW concept. We did not offset these costs in our first trial.

However, we do see substantial data to support the idea that we might see some improvements in Employee Retention over time and that we can significantly reduce the costs of Lost Productivity. We feel the benefits to Employee Engagement and Mental Health combined with the fact that the summer period has a minimal seasonal impact on the business, merit a continuation of the 4DWW program in 2023.

Poll Everywhere was approximately half remote employees until the pandemic and we didn’t fully settle into a remote culture until we finalized letting our office go earlier this year. Part of that transition demands training to improve communication and productivity in our remote working culture. These efforts can substantially overlap with a refreshed experiment in working four days a week more efficiently. We are excited to see our progress in the next trail!