- Determining the goal of the survey;
- Selecting the questionnaire applicable to the goal of the survey;
- Determining the population and sample of the survey;
- Setting the timing of the survey;
- Determining the medium of the survey;
- Gathering contact information of the sample;
- Implementing the instrument used to perform the survey;
- Conducting the survey;
- Analyzing the results of the survey;
- Determining actions for improvement and returning those as feedback.
Determining the goal of the survey
An employee engagement survey or employee satisfaction survey (from now on: a survey) can be conducted in order to determine (measuring) what the current state of engagement or satisfaction is within the organisation. A survey can also be conducted to bring the engagement or satisfaction to a higher level (improving). It makes sense to determine beforehand what the goal will be of the survey. If the goal is to improve engagement or satisfaction, then measuring can be still in the picture. However, that doesn’t work the other way around and you need to have at least a small indication of what needs to be improved in the organisation.
Selecting the questionnaire applicable to the goal of the survey
A survey conducted to only measuring can benefit from a different questionnaire then a survey being conducted to improve employee engagement or to improve employee satisfaction. When a survey is being conducted to determine the current status (measuring), it is important to choose a compact questionnaire of a quantitative nature; that makes the statistical analysis in the end more easy and enables, thus, the organisation to determine the current state with less effort.
When improving the engagement or improving the employee satisfaction within the organisation is the goal of the research, it makes sense to estimate beforehand which areas might need deepening. If there is, for example, a presumption that the working culture is not all right, then it might payoff to add some deepening questions that focus that specific topic. That’s why the advice of using a compact questionnaire doesn’t hold when improving is the goal. Besides that it might payoff to, when improving employee engagement is the goal, take qualitative aspects into account in the survey. These qualitative aspects are mostly best measured by open questions. An example of such a question: “What can we improve in the organisation to bring the engagement to a higher level?”. Answers given on these type of questions give often hints when defining actions for improvement.
Determining the population and sample of the survey
The population of a survey will be equal to all employees of the company in most cases. One might think, however, of surveys in which research is being conducted to measure the satisfaction or engagement within a specific department or within a specific business unit. The same might hold for a specific survey for improving the engagement or satisfaction within one departement or within one business unit. In that case, the population holds all employees within that specific business unit or within that specific department.
Determining the sample can be done manually or can be done, within some applications, automatically. Determining the sample in these surveys is mostly easy: usually the whole population is taken as the sample. This comes with advantages and disadvantages. An advantage of using the complete population as the sample is the feeling among employees and management that everybody is being incorporated within the survey. A disadvantage is however that many participants don’t feel the need to participate every x weeks within the survey. The possibility to have always actual data is thereby eliminated. By drawing a sample by the size of, for example, a twelfth of the population, one should be able to have an actual overview of the satisfaction or engagement within the organisation or within a department.
Setting the timing of the survey
In most cases, the survey has been planned for a long time already; forthcoming out of plans written in the previous year or because the need to improve is visible for a longer period of time already. This means, in most cases, that the survey doesn’t have to be conducted necessarily on date Y. This gives room for choosing the right timing of the survey. Different factors play their role in choosing the right timing: workload, structural changes and presence.
Choose for conducting the survey a period of 2 to 3 weeks in which the following conditions hold: there is no above-average workload, publishing or implementing structural changes has not been scheduled and absence is not at an above-average level.
An above average workload could possibly take out any possibility to participate for those employees involved or could take away attention during participating causing the results to be unreliable. The publishing of structural changes within the organisation or within departments or implementing these structural changes normally comes with a high level of unrest. Besides, the chance might exists that one part of the sample participates before the announcements or implementation and the remainder afterwards. This might as well lead to unreliable outcomes. An above-average level of absence might lead to, which is no surprise, a lower response rate. Do not choose a period of holidays and, more specific, start on a working day at which absence caused by parttime jobs is at its lowest.
Determining the medium of the survey
A survey can be conducted online or on paper. A survey that is conducted online can still have invitations on paper. Sending out the questionnaire or the invitation by mail (on paper) mostly increases awareness among employees and leads to, in most cases, a higher response rate. The same can be reached in some cases by only sending the invitation by paper. Besides increasing awareness, sending the invitation by mail, especially when using the home-address of the employee, might increase the feeling of anonymity and confidentiality which might increase both reliability and the response rate.
A survey that is conducted completely on paper causes in most cases, when it comes to processing the response, extra energy and thus extra costs. The question is whether these extra costs balance with the extra awareness. In some cases, not all e-mail addresses of the employees might be available which then becomes a valid motivation for sending out at least the invitation by mail.
The timing of the survey might be influenced by the choice of the medium used. This holds even more if paper is used as the medium since it might make sense in some cases to send out the questionnaire or invitation, for example, together with the payroll-check or with the companies’ magazine.
Gathering contact information of the sample
The gathering of contact information about the sample (the employees who will participate in the survey) is an activity that is mostly, in terms of energy, underestimated. In those cases where the choice has been made to invite employees by sending invitations by e-mail to their private e-mail addresses, it often appears to be hard to have the availability over up-to-date addresses of all employees in the sample. It might be wise, to prevent having all effort by one individual, to share this activity with the managers. One might decide to send out an empty spreadsheet to managers with the question to return it with all employees’ data within seven days. Some tools offer the option to have it entered by managers of the departments by themselves. This comes often with the advantage that the data entered is directly validated (e.g. whether the postal code or e-mail address is correct).
If it appears to be too difficult, during this process, to gather, for example, all e-mail addresses of all employees then it might make sense to reconsider the choice for the medium for the survey (previous step).
Implementing the instrument used to perform the survey
Several instruments are available to conduct a survey and to ease the whole process. One element of this step might be the selection of such an instrument (software). When selecting such an instrument, please keep in mind the level of automation that is expected. If there is expertise and knowledge within the company to conduct the analyses on the results, then the selection of a software-package for statistical analyses (such as SPSS or Excel) might be enough. Software-packages are available (for example: Jobber) that completely automate the process of conducting and analyzing the survey and its results.
When the choice is not being made in favor of a software-package that does the whole process, then there should be put effort in giving meaning to the medium chosen before for answering the survey. For conducting a questionnaire on paper, a text editor might do the trick (for example: Word). When the questionnaire is being posted online, it might be wise to select one of the tools available for setting up a questionnaire (such as SurveyMonkey).
Determine when selecting an instrument, if the instrument offers the option needed related to the requirements of the sample, the requirements in terms of medium used and the requirements related to the questionnaire. Implementing the software-package doesn’t come with one description since this depends to a large extend to the package chosen.
Conducting the survey
Which activities are needed in this step solely depends on the solution that has been chosen in the previous step to conduct the survey with. When an all-in-one solution is chosen, it might be enough to only activate the survey at the right moment in time. When the choice has been made to go for a lower level of automation, ‘conducting the survey’ might include at least sending out the invitations, gathering and processing the response and, when needed, sending out reminders for participation.
It might make sense, to increase awareness, to emphasize participation in the survey via other media. When there is the availability over displays within the working environment or an intranet, extra attention could be realized by putting a message on it when starting the survey. Besides that, the use of other media might create extra attention when sending out reminders for participation. When the initial invitation has, for example, been sent by e-mail, one might choose to send out the reminder for participation by SMS or by mail. Off course, this is only possible when the related contact information is available.
Indicate in the invitation how and when actions for improvement and / or results will be shared. This not only gives the employee the feeling that he or she is taken serious but creates the obligation for the team conducting the survey or management to put out actions based on the results gathered and to share these with the employees. A survey that has been conducted seriously might increase the engagement of employees by itself.
Analyzing the results of the survey
Which activities are associated with analyzing the results of the survey depends, again, on the software used for conducting the survey. When the choice has been made for an all-in-one solution, results can often be gathered from a dashboard or report. When the choice has been made to perform the analyses manually, analyzing requires a lot more attention and effort.
When conducting the analyses manually, the help of a statistical software package, such as SPSS, is not only helpful but might be required as well. Mostly, only these type of software-packages are able to conduct an analyses on correlations between variables (ANOVA) and to conduct an analysis on reliability of the construct (indicator). Please refer to statistical manuals in order to perform the analyses needed and to know which steps have to be taken in advance when conducting a conclusion on the results gathered.
Determining actions for improvement and returning those as feedback
Determining actions for improvement can be read as determining the necessity of a follow-up study for improvement when only measuring was the goal of this survey. When improving the engagement or satisfaction of the employees was one of the goals of this study, it is important to come to the point where conclusions are drawn from the survey and where actions for improvement are being formulated on the base of these conclusions. Furthermore, these actions for improvement should be shared with the employees that participated in the research again.
It is highly advisable to discuss actions for improvement within the organisation; those actions gather support when these are the outcome of a feeling of agreement. These discussions might be supported by several different instruments (applications) available: a distinction can be made between having the discussion completely held online and anonymous and planning one moment in time in which everyone comes together to hold a session in which these action are being discussed.