From a fuzzy idea to a survey to actionable intelligence: How to plan an employee survey to encourage organizational change.
Written By David Chaudron, PHD
David shares his 11 pointers that allow movement from the “Yeah, we need to find out what our employees are thinking” to specific actions based upon the data collected.
- Create and communicate clear, specific actions from the employee survey data.
Suggesting that “management communicate more” or “we need team spirit” doesn’t do much. What really needs to change? It is also very easy to throw some communication training at supervision, hoping this will paper-over management’s unwillingness to tell hard truths. Training someone implies that lack of skill is the cause of your problems. What if the cause is a systemic issue instead?
- Include the survey process into the normal business planning cycle.
Syncing the schedule of the survey with the normal budgeting cycle increases the chances that recommendations will be funded. For example, if budgets are due in November, and next-year’s objectives are due in October, develop recommendations in September, and conduct your survey in the Summer.
- Don’t try to “game” timing of the survey. I’ve had clients suggest that they don’t want to do a survey now, because they want to announce something “good” right before employees take it.
- First use numerical surveys, then follow with focus groups.
Using focus groups first allows “squeaky wheels” to have too much influence. Allow input from all employees to prioritize issues, then use focus groups to gather richer detail.
- Avoid using agree-disagree scales.
Agree-disagree scales, while commonly used, have response-bias issues, and most importantly, are difficult to interpret, even with norms. To give a quick example: How can you prioritize survey items where one shows 37% agree, but 42% disagree, with another item that is 22% strongly agree and 17% agree?
- Don’t look for what you already see.
Conducting a training needs survey assumes lack of skill is the cause of company problems. Conducting a wide-ranging survey at the start will help avoid agreeing with what you already believe.
- Use multiple survey methods.
No one method (numerical surveys, open-ended questions, focus groups, etc.) is the gold standard of data collection. Each has its advantages and disadvantages.
- Keep the data anonymous, but communicate the actions.
Some employees may be to paranoid about tracing their data back to their computers, we’ve had to revert to paper surveys for some of our clients.
- Decide how to analyze data before you gather it.
How will your graphs and reports look? If they look a certain way, how will you interpret them?
- Decide on your sampling plan, and how to “break out” the data.
Deciding whether to do a 100% sample of employees, or a random sample, is an important statistical (and buy-in) question to ask. Asking too many questions, like gender, location, job title etc. can violate anonymity or the perception of it.
- Involve influential employees in the survey effort.
We involve key employees in the planning effort of the survey. They can become mighty advocates of survey recommendations.
- Never survey without acting.
Even if management decides they cannot (or will not) solve a problem employees raise, it is still important to acknowledge the problem and state clearly why management is not taking action at this time.