Let’s imagine that your Support Desk manager has come to you and asked you to build some Customer Satisfaction surveys (nSAT) in Service Now and you are not sure where to start. Here are some basics to help ensure that you are building effective surveys that provide you and your management team with tangible results, while still ensuring surveys are relevant and utilized by your customers.
Survey Basics – Transactional vs Relational
We will start with the difference between Transactional vs. Relational Surveys. Most of the time you’ll be using a transactional survey, which is a survey that is sent in regards to a specific transaction (ticket) that was submitted to your Support Desk.
Relational surveys are focused on the overall experience of the Support Desk. These are often sent with a substantial list of questions that allow for free-form feedback from your customers. The survey is intended to measure how the experience evolves over a period of time has passed and requires far more in-depth analysis than the transactional survey, due to its nuanced questions and amount of unstructured feedback from your customers.
For this discussion, we are focusing on transactional surveys.
Know your metrics
Before designing your survey, you should know whom you are reporting for, what metrics you are trying to provide data for and how this information will be calculated, as well as when these metrics will be provided for analysis.
- Who – Will this nSAT reporting be for Support Desk management or is this for executive management to evaluate a recent tool change or product release?
- This is important because these stakeholders should know what decisions they need to make, based on the metrics. That will help you know where you need to end up.
- What – Are the stakeholders reporting against contractual service level agreements (SLAs)? Or, is the nSAT to help calculate Support Desk agents’ performance and bonuses?
- If there are existing metrics that you are trying to design a survey around, this will help to clarify the survey questions (i.e., contractual). On the other hand, Support Desk agent metrics could change annually and be more nuanced.
- How – Will your metrics be calculated metrics based on survey questions and measured responses or will you be analyzing free-form feedback from the customer?
- As mentioned, free-form feedback takes additional resources and is often considered “optional” in surveys, while scale-based questions will provide an easier way to analyze data patterns for quick reference to metrics.
- When – This one isn’t as important as the other components, but knowing how often nSAT will be reviewed is important to help calibrate how often the surveys are sent.
Start Where You Are Now
If you are working on revamping your existing surveys, a good place to start is evaluating what survey questions have provided the most meaningful data and what has been the most difficult to analyze. This seems fairly straightforward, but “if it ain’t broke, don’t fix it!” Your customers are accustomed to surveys that have a certain look, feel and length. If you are trying to add in several additional questions, change the format or generally trying to revamp the survey, this could produce false responses, because the customer is used to seeing a survey a certain way and may not take the time to review the new format.
An example of this: Changing the satisfaction emojis (discussed below) to be in a different order. The end user will most likely choose the one to the far left, which is “always” been very satisfied, but is now very dis-satisfied, creating a false result and erroneous reporting.
Work with what’s been working and improve upon it. If you receive the most survey responses from your satisfaction scale of 1-5, but not the open-ended question about how your service can be improved. Keep the scale and evaluate the question. Customers will use the quickest opportunity to complete the survey and leave their feedback. If you want more in-depth responses, consider a relational survey and simplify your transactional survey.
Keep It Simple
Your survey should take no longer than two-three minutes to complete. If there is more than one page, be sure to number the pages of the survey so the customer knows how much more they need to complete before they can submit.
List the most critical questions that provide the most impactful metrics first. And if the customer decides to exit from the survey early, this initial information is saved and submitted. This will help to ensure that if the customer decides not to complete the full survey, you have at least collected the most critical metrics for your analysis.
Try not to include more than five questions total in your transactional survey. Most customers are not willing to spend the time providing in-depth feedback. They either do not have the time, they do not have the patience or, and this is important, they felt that the transaction met their requirements and do not feel the need to provide feedback. This last point is significant because the majority of customers only prefer to leave feedback when something is wrong with their experience and typically don’t when things are okay or great. You will want to capture this information from the customer without frustrating them with lengthy surveys and feedback forms.
Yes, emojis are fun, but in this case they are very helpful. Emojis allow for a quick reference object that the customer can select to register their satisfaction regarding their experience.Emojis transcend language barriers and allow for an instant register of the current satisfaction scale aligned with the appropriate questions. Also, this helps to keep the scale down to a maximum of five options. Any more than this and your survey runs the risk of providing inconsequential data due the results being too nuanced and statistically irrelevant.
→Pro Tip: If your survey is sent out globally, you will want to consider removing the color coding, as color associations are culture-sensitive.
Don’t Burden Your Users
Another variant to consider, is how often surveys are sent. In many organizations, there are power users, who submit tickets to the Support Desk for other users. This individual can submit between five and 50 tickets a month and could receive just as many survey questionnaires.
There are several different options for survey frequencies, but we will cover three for ServiceNow.
- Periodic – ServiceNow allows you to set up the survey submission, where a survey is only sent to a customer, up to once a month. This will help to reduce the number of surveys sent to customers, while still ensuring the occasional ticket submitter receives a survey in a timely fashion.
- Transactional – Set up the survey submission, based on the number of tickets submitted. This could be set up as sending a survey every five tickets only, which provides a randomized process for survey submission, again without burdening the customers with endless surveys.
- User vs. Ticket Based – Both of the options (above), can be set based on the user or the ticket frequency. For instance, you can have surveys sent every fifth ticket, per user or only once a month, per user. OR, you can have the survey sent every fifth ticket based on the service (i.e., INC vs REQ) or only once a day, per service.
No matter what the variation of user vs ticket, time vs transaction-based survey occurrences, this will help to reduce the survey “noise” and improve the chance of customers providing valuable feedback. Remember, survey response rates are commonly below 10% overall so each and every survey you receive is important for developing meaningful metrics.
Reporting and Review
Before deploying your surveys, make sure that you are using all of the native functionality in ServiceNow to provide consistent reporting for your metrics. After all, what is the point of sending out surveys, if you can’t read the results. The suggestion here is to set up weekly reporting reviews for the nSAT survey results. This will help to ensure that your results are being utilized as well as understand who is returning surveys vs. who is not responding.
Lastly, it’s wise to set up periodic review of the survey structure to ensure you are receiving meaningful data from customers. Some other things consider:
- If you are finding that your survey response rates are low, consider adjusting your frequency.
- If the results are only coming back with high level nSAT data and your open-ended questions are not being completed, consider reducing or removing the free-form question and/or changing to relational surveys on a semi-annual basis.
- If there are a noticeable amount of false results, meaning, when a negative nSAT follow up discussion with the customer results in the customer saying that their submission was a mistake and they were actually happy, this is the time to evaluate your question format.
- If you recently made changes to the survey, it’s good to communicate to customers so they can make appropriate adjustments to their behavior.
Overall, survey design and management can appear to be a straightforward approach, but there are some important considerations to take into account to ensure that you are getting the most out of your survey analysis.
Coming soon! Don’t miss our follow up blog: The Steps to Build Out Effective Surveys in ServiceNow
If you would like Covestic’s help with creating surveys in ServiceNow, contact your account executive.