There are numerous ways to slice and dice an organization’s employee engagement survey results. Organizations might look at department, position level, generation, location, tenure, gender, or any other employee demographics that comes to mind. Though these variables can lead to a highly focused and detailed understanding of employee perceptions, it can be easy to get lost in the data.
So what’s a nice “snapshot result” that can be quick to calculate, yet offer a compelling story?
Meet the consensus differential.
Let’s start by breaking down the term: “consensus” means agreement, and “differential” means range.
As a single term, the consensus differential refers to the level of agreement, or range of perceptions, across comparable employee demographics. The consensus differential reveals how similar group perceptions are at an overall level. In other words, it helps you identify potential problem areas.
The chart above shows employee favorability over the last two years at Awesome Co., sliced by location.
To identify the consensus differential for the current year, we simply calculate the difference between the groups with the highest and lowest favorability. Location 1 has the highest favorability at 85.0 percent, and Location 3 has the lowest with 75.0 percent. When we subtract the favorability of Location 3 from Location 1, we get a consensus differential of 10 (85-75=10).
We can also see that the differential increased from the previous year. Last year, the differential was 4.0; Location 1 (82) – Location 3 (78) = Consensus Differential (4). Location 1 increased in favorability and Location 3 decreased in favorability from the previous year to the current year, so the gap between the highest and lowest percentages widened. From an overall perspective, this indicates that the three locations were more perceptually aligned the previous year than they are in the current year. Conversely, the locations are more perceptually divided when comparing this year to last.
This type of analysis can spark all sorts of questions that warrant further analysis. For example: Why did Location 3 decrease? Why did Location 1 increase? Why did Location 2 remain the same? Do employees from Location 1 and 3 ever interact? If so, do their differences in perceptions seem to affect collaboration and communication? If customers walk into Location 1 versus Location 3, do they get a totally different experience? If so, how does that influence location-specific sales, referrals, and client retention? And lots, lots more.
When determining whether a consensus differential is at, above, or below average, two variables must be taken into account: the type of employee demographic, and the number of groups compared.
First, decide the type of employee demographic you want to analyze: functional (e.g., business unit, department, division), geographical (e.g., location, country, region), and hierarchical (e.g., position level, managerial status).
Then, determine how many groups you want to examine: four locations, five departments, etc.
Lastly, multiply the number of groups with the “rule of thumb” demographic multiplier in the chart below to calculate the typical consensus differential for this group.*
| Rule of Thumb Multipliers | |
| Employee Demographic | Multiplier |
| Functional | 2 |
| Geographical | 2 |
| Hierarchical | 3 |
For an example, let’s look at the above graph again. Location is a geographical demographic variable, and Awesome Co. has three locations (creatively named Location 1, 2, and 3). Because the geographical multiplier is 2, the typical consensus differential for this group would be 6 (2 x 3 = 6).
With that in mind, we can see that the Previous Year Consensus Differential of 4.0 is below that benchmark of 6, meaning the locations had lower-than-average perceptual differences. In other words, they’re more aligned than what we typically see for that number of geographical groups because 4 < 6.
On the other hand, the Current Year Consensus Differential of 10.0 is above that benchmark of 6, meaning the locations have higher-than-average perceptual differences, suggesting that something occurred within or across locations to make them less aligned (10 > 6).
One caveat is that those ratio rules tend to be less reliable as the number of groups increases, such as when the number of functions exceeds 20, the number of locations exceeds 15, and the number of positions exceeds 10. But even if your organization is highly segmented, geographically dispersed, or especially layered, these rules of thumb can give a good approximation of typical consensus differentials.
Before concluding, I want to note that as an outside observer, I can’t give value judgments about higher or lower consensus differentials. A lower differential may be preferred across locations to ensure perceptual consistency from one store or office to the next. A higher differential may be preferred across position levels, with executives and senior leadership having much higher levels of favorability than supervisors and individual contributors. Whether a consensus differential is too low, too high, or just right depends on your organization’s culture, strategic objectives, and the ways in which your organization reacts to engagement survey results.
Ultimately, the consensus differential is just a range of perceptions, one number that can be easily calculated when looking through employee engagement survey results. Behind its simplicity lies interesting, complex, and actionable stories for further investigation. And with the “rule of thumb” multipliers listed above, you can easily determine how the range of your employees’ perceptions compare to other organizations.
The sample used in this study is comprised of 67 randomly selected organizations that have taken Quantum Workplace’s engagement survey. Weighted averages were calculated for each distribution type so as to accommodate for size differences across groups.
Looking for more ways to understand, analyze, and communicate employee engagement survey results? We’re always publishing best practices on our blog. Click below to receive new articles right in your inbox.
Published March 2, 2017 | Written By Dan Harris, Ph.D.
