Five Tips for Improving Response Rates

When making plans to create and distribute your survey, here are five things to keep in mind.

1. Use Incentives - For surveys that are not compulsory for the respondent to complete, it's common sense that providing incentives generally results in better response rates. As a rule of thumb, research indicates that the larger the incentive, the larger the response rate. Consider offering a raffle drawing prize or perhaps an incentive for the first 50 respondents. Managing your incentives in this way also means that you don't have to spend a fortune on incentive rewards. Also, it is important for the respondent to actually believe that your survey is credible and that the incentive reward will actually be given. Providing instructions to the respondent that include an email address with your organization's email domain name (i.e., yourcompany.com) or some other means of contact (a phone number or mailing address) can often provide the necessary impression that your survey is legitimate. 

2. Include a Direct Appeal for Participation - It's subtle, but an important point that can easily be overlooked is how you appeal to the respondent to participate in your survey. The cover letter or email that accompanies with your survey questionnaire can have a big influence in response rates. Specifically, research indicates that it is a good idea to include a sentence or two specifically asking for the respondent's participation citing the importance of their helpful input in achieving the objectives of the survey (i.e., improving a product, process, work environment, etc.). Doing so appeals to the respondent's sense of being a helpful contributor when called upon to participate.

3. Clearly State the Purpose - Let's face it, if a respondent can't figure out in a few sentences what the objective of your survey is and how they can contribute, then they are less likely to respond. Being forthright in your cover letter and initial instructions as to why you are conducting the survey and how the results will be used is important. 

4. Set Time Expectations - Surveys that take less time to complete typically result in better response rates. Review the objectives of your survey and keep only those questions that are essential in achieving your objective. Including a statement in your cover letter or initial instructions that the survey should take, for example, no more than five minutes to complete can positively influence a respondent to participate. You will want to, of course, have several people take the survey before widely distributing it to make sure that your claim of five minutes is accurate. 

5. Fine Tune Your Survey by Getting Feedback in Advance of Wide Distribution  - It's a good idea to distribute your survey to a small set of people (preferably accessible colleagues or constituents) who actually complete the survey before you send it out to a broader audience. You can then follow-up with these respondents via a phone call or email asking them if their experience with the survey was difficult or confusing in any area. This provides an opportunity to collect feedback from actual respondents and make fine tuning adjustments to any confusing or hard parts in the survey. Sometimes rewriting a question or adding specific choices makes all the difference. This vetting process produces dividends because you can have a better confidence in your survey when you send it out to a broader set of respondents.

Three Tips for Working with Survey Results

Here are a few tips for working with the survey results collected in your SurveyGold software.

1. For a Long-Running Survey Use a Date Range Filter Rather than Creating a New Survey - One of the most common inquiries I receive pertains to how to manage long-running surveys. Specifically, many SurveyGold users are conducted assessments surveys over time so that can measure, for example, last year's results with this year's results. 

The solution that many SurveyGold users employ is to use the Copy Survey command in the Setup Survey tab to create a copy of last year's survey and just change the name slightly so that it becomes  this year's survey. However, my recommendation is that you do not create a new survey, but continue to collect responses using the existing survey. Then you can use the date range filter feature in the View Results tab to view only the respondents from a particular time period such as last year. The benefit of this is that you can easily create a filter for last year and another filter for this year and then compare the two sets of respondents as multiple series in a single graph. This is not possible if you create two separate surveys and is only possible by using the date range filter from the entire set of long-running survey results.

Click here to watch a tutorial that shows the filter and multiple series features in action >> 

2. Use a Chart Title to Eliminate Verbose Question Text in a Graph or Report - It is often necessary to write a lengthy question in order to convey the intent of the question to the respondent. Later after the results have been collected and you begin to view results in graphs and/or reports, the lengthy question text results in visual clutter in those graphs and reports.

The solution to this problem is not go back and change the question text from verbose to succinct. Rather, in this case I recommend that you use the chart title feature of SurveyGold that provides the ability for you to specify a more succinct chart title without changing the question text itself. 

You can change the chart title associated with a question in one of two ways:

	In Setup Survey, via the Chart Title tab in the question properties sheet
	In View Results, by right-clicking while viewing a particular question and selecting the Change Chart Title popup menu item

3. Keep Question Response Choices as Terse as Possible - Keep in mind that any SurveyGold survey question that has choices (in other words, it is not a fill-in-the-blank question) appears as a graph in the View Results and Analyze Results graphs and reports. So, it helps to keep question response choices as terse as possible so as to avoid crowded or truncated graph labels and legends. 

Even if you have a situation where you cannot avoid creating lengthy question response choices, it is possible for you to influence the way that the choices appear in a graph:

1. From the SurveyGold main menu select the View Results tab.

2. Select your survey from the list.

3. Select a question that you know has lengthy question response choices and click View Results.

4. When the pie graph appears, right-click and select "Change Legend Font Size". When the dialog appears, change the legend font size from Auto to 8 and click OK.

5. When the pie graph appears, right-click and select "Change Label Font Size". When the dialog appears, change the legend font size from Auto to 8 and click OK.

This makes the legend and label fonts as small as possible which makes the text that appears in the graph which makes the legends and text as small as possible thereby avoiding truncation of the response text or legend text in the graph.

Chuck Boudreau is the Author of SurveyGold Software. In his spare time, Chuck and his wife and kids enjoy dog training their Hungarian Puli named Kashi.