Survey Monkey has now been in existence for over 10 years and has allowed a wide range of people and organisations to carry out market research they might not have otherwise been able to afford. Although the software definitely has its limitations I have successfully used it for relatively simple surveys up to around 20 questions in length.
Below are 10 tips for getting the best out of Survey Monkey in terms of making your surveys as respondent friendly as possible. The more your surveys are designed with respondents in mind, the higher the rate and better the quality of response.
1. Subscribe to the Professional version of Survey Monkey. It costs a bargain $19.95 a month (less than £15 at current exchange rates) and, amongst other features, allows you to include skip logic in your questionnaires, add a logo to the survey template and cross-tabulate the resulting data. In any case the Basic version only allows a maximum of 10 questions and 100 responses so any serious research will need the paid for version. You can subscribe on a month by month basis so you only pay when you are using the software. (By the way, I have no connection with Survey Monkey and, unfortunately, do not get any commission on any sales I may generate!).
2. Improve the look of your questionnaire. It never fails to amaze me that large companies who have all their communications professionally designed with a consistent feel are happy to send out questionnaires designed in Survey Monkey that look awful and reflect badly on their organisation. Questionnaires should be as clear, uncluttered and as easy to read as possible.
Although you don’t have full flexibility to make the questionnaire look exactly as you want within Survey Monkey you can edit the themes available to you if you subscribe to the Professional version. You can change a number of aspects of the way your survey looks including the font size and background colour. You can also add your company logo to the top of each page. If you have some basic HTML knowledge you can also format text within a question, for example underlining or emboldening certain words within a question.
You can find the “Edit themes” button if you click on “Edit Survey” when within the “Design Survey” area.
3. Split your questionnaire into different pages. I see a large number of Survey Monkey questionnaires where the survey is presented on one page and the respondent has to scroll down to continue with the questionnaire. As well as putting the respondent off, answers to questions may be influenced by respondents looking down the survey at future questions. Ensure that the questions on each page fit onto a screen so that respondents do not have to scroll down. As a guide no more than 3 questions should be on each page.
When designing your questionnaire you should see a button called “Split Page Here” after each question. Make sure you use it. This feature is available in the Basic version of the software.
4. Use Skip Logic so that respondents only answer questions relevant to them. Something else that infuriates me is a long questionnaire that includes instructions such as “If you have answered No to Q5 then go to Q7”. Respondents should be concentrating on answering the questions not guiding themselves through the survey. These sorts of instructions are likely to confuse respondents causing them to answer irrelevant questions, miss out ones they are supposed to answer or to close down the questionnaire completely.
The logic you can use within Survey Monkey is relatively unsophisticated. Questions, or blocks of questions, can only be filtered from the previous question. Combinations of answers from one or more questions cannot be used as filters at any time.
Splitting the questionnaire into pages means that respondents can skip pages without ever seeing the questions they should not be answering.
You can add skip logic for each question when designing the questionnaire. Skip Logic can only be used by subscribers to the Professional version of the software.
5. For open-ended questions increase the size of the box where respondents type their answers. By default the box that appears when adding a “Single Textbox” or “Comment/Essay box” question is fairly small. Small text boxes encourage respondents to give short, shallow answers.
Clicking on “Edit question” for an individual question presents you with a number of options. Scroll down to the “Change question size and placement” option and check the box next to it. This brings up a number of options. Using these ensure the box is as wide as possible (100 chars wide) and increase the number of lines to a reasonable amount, I would suggest 5 or 6. This feature is available in the Basic version of Survey Monkey.
6. With a few exceptions, make each question a must answer question. If this is not done respondents can skip questions as they wish making the data incomplete. However, it may be the case that you are happy for respondents to skip some questions where they might not have an answer i.e. is there anything else you’d like to tell us about product X?
Clicking on “Edit question” for each question allows you to check the box “Require Answer to Question”. The default is for this box to be unchecked. Again, this feature is available in the Basic version.
7. Include a “Thank You” page at the end. Once the respondent reaches the end of the questionnaire they should reach a page that thanks them for completing the questionnaire and provides details for someone they can contact if they have any queries about the survey. Including a “Thank You” page is good manners and reassures respondents that the survey is a genuine piece of market research.
These contact details should also be provided in any introduction to the survey either in an email invite or at the start of the survey. The introduction should also explain the purpose of the survey in order to maximise response.
To create a “Thank You” page add a new question and choose the question type “Descriptive text”. You can then type text as you wish.
8. Make sure the survey links to a relevant site on completion. By default when respondents finish the survey they land on the Survey Monkey website. Professional subscribers can change this within Survey Monkey to any website of their choice. This is a great chance to guide respondents to a relevant website and, at the very least, increase traffic to that site.
The facility to change the end page link is available in “Change Settings” when in the “Collect Responses” area.
9. Unless there are good reasons not to, allow only one response from an individual PC. This will stop rogue respondents going into a survey a number of times to complete it. Respondents may be tempted to do this if the survey has some sort of incentive e.g. entry into a prize draw, or if an individual respondent cares so much about the subject of the survey that they wish to bias the results in some way. However, if you expect different respondents to complete the survey from the same PC then you should allow multiple responses from one PC.
In “Change Settings” within the “Collect Responses” area there is “Allow multiple responses”. By default this does not allow multiple responses per computer but it is worth double checking this before sending out a survey.
10. Test, test and test the survey repeatedly. Mistakes in questionnaires reflect badly on the organisation associated with the survey and impact negatively on response rates.
Go through the survey many times yourself before allowing real respondents to access it to check everything makes sense and is working as it should do. You will need to temporarily allow multiple responses from one PC to do this. Also, get others to go through the survey, preferably, those for whom the survey is relevant e.g. if you are sending a survey out to football fans, get a handful of football fans to go through it before you send it out.
Once you are happy with the survey make sure you clear any data you have collected through testing (an “eraser” button is shown next to all of your surveys when you are in the “My Surveys” area. Click on this button to clear data for a particular survey). Also, make sure that multiple responses are not allowed from the same PC before sending.
You can access a short , simple example survey which has incorporated these tips by clicking here. This survey does allow multiple responses from one PC so you can go into it as many times as you like.
It should be remembered that these tips are specifically related to Survey Monkey and do not cover the specialist areas of sampling, questionnaire design and data analysis. If you do not have expertise in these areas then it is unlikely that any survey you design will meet your objectives and probable that resulting data will be useless or even dangerously misleading. If you do not have market research expertise but budgets are tight consider hiring a market research consultant who can help design your survey (perhaps someone like myself - details at www.austinresearch.co.uk!).