When designing and setting up an employee engagement survey, the overriding consideration is to ensure the utilization of good science. But, it’s not just about good science. Even more critically, the process needs to take full advantage of the right science. The biggest mistake and biggest factor that has limited the impact and potential of employee surveys over the years is that the wrong science was applied to the survey process.
In way too many instances, the individuals responsible for designing the process, with the best intentions, applied techniques that may be legitimate for some types of research, but not others. Specifically, they relied on what they learned from conducting successful customer or consumer research, or what they learned in grad school about survey research in general. However, in doing so, they were missing an in-depth understanding of the critical ways employee research differs from customer or consumer studies.
Here’s the epiphany that it took many years of practice and experience to discover, document, and articulate. The requirements, expectations, and influencing factors for people who are buying something, differ dramatically from those same factors for people who are being paid for delivering something. Sellers behave, expect, compromise, and commit differently than buyers. Therefore, the science and the techniques that work for one are most likely to be inadequate, or even misleading, when applied to the other.
To illustrate, when you are paying for something you expect perfection. You may love a new brand of shampoo. You love the way it lathers, the way it makes your hair feel soft and clean, the packaging, the price, and even its advertising, but if you don’t like the way the shampoo smells, you’ll keep shopping and buy another brand. Why? Because you want it all.
Work, however, is called work for a reason. If you want your paycheck, as well as the recognition, respect, and personal gratification that comes from a successful career, you have to make some sacrifices and compromises. You know you have to give some things up to achieve what you want. The same is true for a successful employee engagement design. Good employee research that utilizes the right science, can help employers learn the factors of engagement that are the most critical, manageable, realistic, and affordable. By making improvements and focusing on a small number (1-3) of these important factors at a time, companies can build trust and commitment and actually earn forgiveness for the workplace factors that aren’t as ideal. This is how an engaging workforce gets it done. It may not be perfect but it still keeps us motivated and engaged.
Following are six best practices contrasted with seven of the most common types of mistakes made in Employee Engagement Survey projects. We call them powerful practices and proven pitfalls.
Except for open-ended comments, rely on a five-point Likert agreement scale that always includes a midpoint. This is the scale that is used in most legitimate normative data bases. This scale is also more comfortable to employees because it accepts a mid-point of “neither agree, nor disagree” or “neutral or undecided” as legitimate and acceptable. This is not only correct, but also an important thing to do in employee engagement research. For example, if you want to get a quick win, then the ability to identify where many employees are on-the-fence can sometimes be very important.
Relying on a four or six-point scale with no midpoint forces respondents to make a stand. This strategy is sometimes used in consumer research because you need a decisive response. However, in the workforce, forcing a response is inappropriate because indecision is a legitimate reaction If individuals are legitimately undecided or on-the-fence they become frustrated when their perceptions or attitudes are not included as options. Some will skip the item, respond randomly, or even exit the survey since it doesn’t include an option that accurately reflects their honest reaction. In any of these examples those employees’ data is lost and so is the value of their feedback. Four-point scales are highly appropriate when you are trying to force a direction or choice. It’s done in consumer surveys all the time, but, remember, you’re not voting or buying shampoo, you are trying to understand your people better. Being on the fence is not only legitimate, it provides valuable feedback.
Use an index that includes at least three or four items to ensure reliability. The index should be validated as the best overall predictor of the most performance indicators. The items should be general-summary types of items that do not share meaning with any of the other specific items on the survey—thus biasing the driver analysis.
Use extreme wording when appropriate. For example, satisfaction is a minimalist response that could be more about how the employees think. You want to achieve extreme satisfaction which is about how they feel, what’s in their hearts. It’s also important the Engagement Index includes an advocate item, a pride item, (especially when a strong brand is involved) and an intent to stay item.
Often, organizations utilize an index that includes items that rate discretionary effort. Discretionary effort is a critically-important important outcome of an engagement program, but it is a terrible measurement. Employees have shown, for many reasons, they simply cannot be objective or accurate when they rate discretionary effort. Think about it, we are all taught to always give 110%, right? So we are trained to think that’s what employers want. Instead of using it as a measurement, use the drivers of your culture that empower discretionary effort, this is when the true magic happens.
Additionally, If you include an item like, “Taking everything into account, this is a great place to work.” you aren’t really measuring anything. What defines “a great place to work?. If you merely want to be a great place to work as your main objective, make it great and fun with things like free food and snacks, spa time and an onsite masseuse, shorter work weeks, longer lunches, lower expectations to reduce stress, and have lots of fun activities in the employee lounge. You’ll keep some employees happy, but your high achievers will go elsewhere because they want to be challenged, and get better definition of themselves and their talents. A “great place to work” is defined differently by all of us, your goal is to measure the things that make your business a great place to work and strengthen them.
For the core of the survey, always include specific, unambiguous, and actionable items from a variety of potentially engaging issues or dimensions. From among the many choices always include the right mix of proven items from the six most likely drivers of global employee engagement. Those drivers of engagement are Future Vision, Leadership Trust, Growth and Development, Recognition, Communication and Involvement and Belonging. If thee six issues are not adequately represented in your survey, the odds of you missing something critically important increases significantly.
Remember, you are trying to identify the key factors that drive engagement, so never include items that are obvious outcomes of engagement as opposed to being the drivers of engagement. You are trying to measure cause, not effect. Example: “I have a best friend at work.” This item really frustrates managers, not because best friends aren’t important, but because managers can’t create them. Engaging work environments facilitate more cooperative, positive relationships, which in turn, facilitates more friendships. Engagement drives friendships. Not the other way around.
Also, avoid items that ask employees to rate themselves as opposed to manager or leader behaviors. Too many people lose objectivity when rating their own behavior. Example: “I know what’s expected of me at work.” Instead consider: “At work, expectations are clearly communicated.” It’s the manager’s job to communicate expectations, not the employees’ job to figure it out.
Other than the engagement-index items, which are more general or summary types of items, make sure that all the core items are clear, specific, and actionable. Simple is better than complex. The meaning of the item should not be in doubt or open to interpretation. Each item should highlight an issue that is actionable and readily influenced by some level of management or leadership. If something can’t be changed or improved, don’t ask. Best case examples: “I regularly receive appropriate recognition when I do a good job.” or “The leadership of this company has communicated a vision of the future that motivates me.”
Always avoid items that are broad enough to be part of the definition of engagement because they interfere with the priorities analysis. Additionally, avoid question items that can be interpreted in multiple ways and have different meanings because managers will have challenges determining what actions to take from them. Example: “I like the kind of work I do.” There are simply too many reasons to like one’s work. To be effective, all core items should be clear, specific, and actionable—and never overly broad.
Also, avoid items about fairness in the workplace. Employees use dramatically different standards when judging fairness. Your best employees will probably feel that, since they do twice as much, they should be paid twice as much and appreciated twice as much. Others are likely to believe it’s only fair when everyone is treated the same. Which group do you want to listen to?
Other than the six most-likely drivers of employee engagement discussed above, thoughtfully consider what other issues might be uniquely important in your particular culture, and therefore merit inclusion in the survey. Some examples that could, at times become factors in making engagement either better or worse, might include issues like diversity, change management, compensation, and benefits, service quality, or work/life balance.
Never let anyone tell you that one-size-fits-all. There are a few select items that are universally important to engagement and merit inclusion on every engagement survey, but there is no single survey that is the best fit for all companies. Every organization has a unique culture, different obstacles to overcome, different opportunities and different risks. Your survey should always be designed to fit not only your company but also your current initiatives and key objectives.
Most successful engagement survey processes include a few items that are not likely to be drivers of engagement but are important for other reasons. These items can vary significantly based on the type of business you are in. They are always designed to measure the factors that will allow engaged employees to be better equipped to be productive and successful in your particular business. These factors could include things like availability of resources, emphasis on customer service, availability of appropriate training, service quality, and safety.
Don’t include more than forty-five or fifty items. After fifty items, some employees lose interest, begin to respond much less thoughtfully or in random patterns, or stop responding altogether. This clouds the research findings.
On the opposite end, never accept the promise that a valid and effective engagement evaluation and diagnosis is possible with just twelve to eighteen items. The majority of highly-successful survey processes include between twenty-eight (28) to thirty-five (35) question items that can be completed in less than five minutes (not counting open-ended comments, which are normally optional, if used at all).
The key points outlined in this paper are based on experience and insight from over forty years of research, experimentation, observation, and study.
When planning and designing your next employee engagement survey, use these powerful practices and proven pitfalls as insights to drive the impact and success of your survey efforts. If you would like additional information on any of our findings, please contact us.