How to Analyze Survey Data

When you plan your online market research study, you still need to remember the standard steps for any market research project, plus some online specific tasks:


The time frame for an online study is, as discussed previously, often shorter because of the immediacy of the results. Keep this in mind but be sure to give yourself enough time to analyze survey responses.

It is crucial that you consider reporting a main element of your web survey design at the outset of your research project.  What you can say about your results hinges heavily on the types of analyses your questions and the capabilities of your response scales.  Today, I will outline the difference between the two major branches of statistical analysis available for most survey data: descriptive and inferential.

Descriptive statistics are the basic measures used to describe survey data.  They consist of summary descriptions of single variables (also called “univariate” analysis) and the associated survey sample.  Examples of descriptive statistics for survey data include frequency and percentage response distributions, measures of central tendency (which include the mean, median and mode), and dispersion measures such as the range and standard deviation, which describe how close the values or responses are to central tendencies.

Inferential statistics offer more powerful analyses to be performed on your online web survey data.  As the names suggests, this branch of statistics is concerned with making larger inferences about social phenomena.  This can include associations between variables, how well your sample represents a larger population, and cause-and-effect relationships.  Some examples of inferential statistics commonly used in survey data analysis are t-tests that compare group averages, analyses of variance, correlation and regression, and advanced techniques such as factor analysis, cluster analysis and multidimensional modeling procedures.

By designing online questionnaires and survey web forms with a good idea of what you want to do with your data after it's collected, you can create cohesive, powerful reports and presentations. Need more tips for how to analyze survey data, read some of these data analysis posts.

Many people ask the question How do I create a customer survey? The basic steps are the same no matter what type of client survey you're writing: customer satisfaction, global market research, product development surveys, etc.

Step 1: Sit down and figure out what the goal of the survey is. Are you trying to identify upsell opportunities? Want to discover features missing from your current product? Figuring out if it's a good idea to take your marketing overseas and attack a global market? In the very beginning of the survey planning process, you should know what it is you want to get out of the consumer survey. If you don't have a firm customer satisfaction survey objectives in the beginning, while you go through the other steps such as writing survey questions or selecting the best survey software, you're going to stray from the path. If you stray from the path, you may find the final survey results are not as helpful as you had hoped.

Step 2: Decide on a research methodology. Your goals should help you on this step as well. You need to first decide if you're planning to do qualitative or quantitative research. From there narrow the scope further, if you want to do qualitative research are you interested in focus groups, advisory boards, one-on-one interviews? With quantitative research you may decide on comment cards, feedback forms and surveys. Is your survey method going to be online, telephone-based or paper questionnaires?

Step 3: Survey Design. I'm making the assumption since you're reading a survey blog about how to create customer surveys, you're not interested in the other market research methods right now so I'm going to focus on the process of building customer surveys. Once you've gotten through the first two steps, you're ready to start writing survey questions (Finally! I bet you thought this would be the first step!). Customer satisfaction survey design can be the biggest challenge. Luckily, there is survey designing software to help you step through this. Survey software tools often have templates and question libraries to help you write good survey questions.

Step 4: Data collection. Okay, you've picked your customer survey methodology, created a client survey and you're ready to field your survey (or use the data collection tool in your survey application to collect responses). Exactly what you do in this step will depend on what type of survey you decided to collected: telephone, paper, online. One way to get survey responses is to use email marketing tools to send personalized email surveys. You can also share the link on your website, social media sites, invoices, etc.

Step 5: Analyze customer feedback. Analyzing survey data is one of people's least favorite parts of the surveying process. We have some tips for how to analyze survey data here. Don't be afraid of this step. You need to conduct the survey customer analysis to achieve your goal. It's what you set out to do, so keep your chin up. You're only a step away from the final product (and once you choose survey analysis methods you should be almost finished).

Step 6: Share the survey findings. This is what you set out to do. Get answers to your customer questions. Take the customer feedback analysis you completed in the last step and format it. You're creating a survey report you can share within your organization (and maybe with others outside of your organization). If you need tips for creating survey reports or an example survey report, you can read more about them here.

Step 6.1: Take action. This is still part of step 6, but it's important enough it should be broken out. In your customer analysis survey report, you should have shared your recommendations for moving forward. Make sure you make recommendations and there is an agreement about moving forward based on the customer survey findings. If you don't plan to take action in Step 1, then you should save yourself the time of conducting the customer research in the first place.

Any other survey research design tips? How have you used these steps to create a customer survey that improved processes in your organization?

A year ago, the Cvent Web Surveys Blog posted it's first post. Over the past year, we've shared many survey best practices with you that can be applied to online surveys and paper questionnaires alike. Here's a list of some of our most popular posts:

• Tips For Providing Survey Incentives: We get asked all the time about incentives and their place in survey research. There are arguements for and against offering survey incentives. This post isn't about those sides. Instead, it focuses on tips for offering survey incentives if that's the path you choose to go down.

• Offering An Additional Comment Section Is An Online Survey Must: Simply asking for customer or employee feedback implies you're ready to make changes, you want to identify problems and fix them. If you fail to offer a "last ditch effort" with and additional comments (optional) question type, you could be making a huge mistake. Depending on your survey design, it may be the only place a survey respondent can give you truly honest, unprompted feedback.

• Survey Question Flow Impacts Survey Findings: Question order is an important part of good survey design. Whether it's a paper survey on customer satisfaction or an online questionnaire to gather employee feedback, how you order your survey questions could impact your survey findings.

• Survey Report Tips: Writing the Executive Summary or Setting the Stage: When writing a survey report, you can't just launch into the survey responses and the data collection results. You need to provide the report reader with some general background information about the survey project including why you ran the survey, what the goals were, what the data collection methods were. Then you can summarize the results and make recommendations.

• Five Steps to Begin Interpreting Online Survey Results: The hardest part of any survey project is creating the questionnaire. If you design a good survey and plan the analysis, interpreting the data should be a walk in the park. Make sure to review these five easy steps for analyzing survey data.

• Analyze Open-Ended Questions Faster with a Quick Trick: We all know using closed questions are easier to analyze. But sometimes they just don't cut it. This post gives you a quick six step process for how to analyze survey data quickly to summarize survey responses without spending days on the survey analysis.

• Difference Between Causation vs. Correlation in Survey Data: To be able to prove causation, you need to be able to rule out all other possible explanations for the connection. This post aims to debunk the myth that correlations found in survey data means that one thing caused the other.

• What To Look For When You Move Your Online Survey To Test Mode: It's definitely an online survey best practice to move your web survey to test mode before sending it out to your survey sample. Make sure to use this 7 step checklist to avoid any online survey pitfalls.

• Market Research Process: 6 Steps to Project Success: When working on survey research projects - be it a customer satisfaction survey or product development survey or a staff performance review form - you should always follow this six step process to make sure you get the maximum return on investment from your survey project.

• Online Survey Question Pitfalls—And How to Avoid Them: This post points out common online survey pitfalls that are in surveys all the time. However, identifying problems with your work performance assessment or customer feedback questions is only half the battle.

I came across this definition of a client survey on CSM that I think is useful:


I might disagree with the fact that a client feedback form is automatically objective, but with the right client survey question creation and survey method it a client survey can be pretty close to completely objective. What the definition does not point out is just because you conduct a client satisfaction survey, decisions wont be made for you. You need to take steps to analyze survey results after the data collection. Customer surveys give you the opportunity to review customer and client feedback and analyze results to get you going down the right decision making path.

So what can conducting client satisfaction questionnaires help you do?


We have some tips for how to analyze survey data - which is the goal of your client satisfaction right? To be able to analyze survey responses and use the data to make decisions.

As I've mentioned plenty of times before, a critical step in the survey research process is taking the time to analyze data (check out these tips on how to analyze survey data) and presenting your survey findings. If you're like most people, the thought of having to create a survey report makes you cringe - perhaps even hide out under your desk until you think everyone has forgotten about the survey research project. Shame on you if you're under your desk! Why did you even bother spending all that time on the market research or employee questionnaire or customer satisfaction marketing survey in the first place if you didn't want to share the findings?

I do understand, despite my sarcasm. Creating a survey report can be daunting. After all, where do you even start? In past posts, I've mentioned how great Cvent presentations and reports are for post data collection survey analysis and report creation. It's awesome that our online survey software can do a lot of the heavy lifting with the graphical representation of your survey, but you still need to structure the report and provide recommendations.

After you finish analyzing survey results, here's a quick guide to what should be included in your survey report:


Even with these survey tips for creating survey reports, you may still not have the extra time to allocate to building reports and presentations. If this is your situation, let our Survey Professional Services Group help. Our talented team created countless custom, professional reports for our clients to share with their management teams and decision makers. Contact us to learn more about our Professional Services.

Charts, graphs and tables are a must in survey reports. When figuring out how to analyze survey data and which survey analysis techniques you want to use, take a moment to consider how you're going to graphically represent the data. It wont influence your survey analysis methods that much, but it's a good idea to know where you're trying to get to when you start the journey. (Figuring out which survey analysis techniques you plan to use should happen at the beginning of the survey project.)

Charts are one of the most important parts of a survey report because they're the easiest way to look at survey results and understand what's going on. Next time you're working on a report, keep these survey report best practices in mind:

• Don't use 3-D charts. When you tilt a chart, while it may look more visually appealing, it's skewing the graphical representation of the data. Look at the two graphs below, can you order the size of the sections?


In the 3-D chart, just like in the 2-D chart, the purple and blue sections represent 20% of the pie. Do they look equal? What about the teal and red sections, are they equal? Yes. It's hard to see that though. The purpose of charts are to make it easy to see survey responses, don't make it harder by tilting the y-axis.

• Start axis at zero. Always start your axes at zero. Just because you're smart enough to realize you changed it, not everyone else looking at the survey report will be.

• Label your chart. This may seem like a no-brainer, but always remember to label your chart including axes and scales. You'd be surprised how often labels are left out, leaving report readers having to comb through the report looking for what the chart is.

• Colors. Charts and graphs brighten up reports, make sure to use colors that contrast enough that it's easy to understand where one category ends and another begins.

• Even though you're providing the reader with a graphical representation of the collected data, you still need to surround the graphics with explanations. Why is the graph significant? What does it prove? What suggestions do you suggest based on these findings?

Many survey software companies provide built in survey analysis software, but sometimes you may still have situations where you need to analyze your survey results with an additional survey analysis tool. No matter what tool you use, whether its the presentation and reports section of your web survey software tool or a report builder in another survey analysis software, keep these survey report tips in mind.

When working on questionnaire design, some people worry about using open text boxes because there's no way to ensure the answers are entered in the format you need them. Take this survey question example, assume the survey project requires anonymous answers:


If someone says yes, you should have follow up questions asking for name, phone number and email address so your team can follow up effectively. But how do you ensure someone enters a complete email address or that the phone number format is consistent? That's where response validation comes in.

Cvent's survey building software allows surveyors to define the format of the open text box one line questions in the question creation process. In the case of a phone number, this prevents you from having to reformat all the phone numbers.


By defining your question format you save yourself time when it comes time complete your data analysis. If you're ready for that step now, check out these tips for how to analyze survey data from open ended questions.

The hardest part of any survey project is creating a solid questionnaire. If you create a good survey and plan the analysis, interpreting the data should be a walk in the park. Here are five easy steps on how to analyze survey data:

1. Take a birds-eye-view. A good first step of interpreting survey data is to take an overall look at the data and responses. Consider these questions:

• How many people completed the survey?
• What was the response rate?
• What was the average amount of time respondents spend completing the questionnaire?
• How well do the respondents represent the survey target group?

Answering these questions at the beginning should help you understand how reliable your data is and evaluate how potential biases could be skewing survey results.

2. Look at the responses to your key survey questions. If the goal of the survey was to evaluate customer satisfaction and identify possible product enhancements, a key survey question may be what percentage of respondents said they were unsatisfied and what percentage were extremely satisfied. What kinds of enhancements could improve the customer experience and increase the satisfaction with the product?

3. Cross-tabulate demographic characteristics with responses to key questions. By cross-tabulating the responses by demographic characteristics such as age, gender, experience levels will help you which features may be more important to key groups. Cross-tabulating helps marketing executives further identify target markets.

4. Look at the open ended questions. We've given you tips to help analyze open ended questions in the past. Using open ended questions in any type of market research or customer survey should help you identify the language your target market is using. Sometimes marketers think customers understand the messaging they choose, while the customers are describing the problem and the ideal solution in a completely different way. Open ended questions can help identify some of these disparities so the marketing department can create messaging to speak your customer's language.

5. Create a "to do" list and take action. We've mentioned before that conducting surveys, whether they're online questionnaires or paper feedback forms, create expectations. Be sure to act on the survey results.

Giving the survey analysis techniques as much thought as the survey creation is critical if you want to be able to really improve your organizations value adds. Some people believe that the analysis process beings only after the web based survey has been launched and responses have been collected. Unfortunately, this approach will hurt your survey results. You should consider the types of customer analysis you want to do while developing a questionnaire. The question type you chose when creating an online survey will limit the analysis. So while the above steps have to do with interpreting the data, you need to think about the process in the beginning as well.

We've stressed the importance of careful consideration when it comes using to open-ended questions in your surveys, as analyzing their responses can be a real hassle. If we didn't warn you early enough against the post-collection headache of open-ended questions, or you decided open-ended questions were the way to go but now you have more responses than you can wrap your head around, here's how to analyze survey data quickly to summarize responses without spending days on the project. We know you may be shocked that it's possible to analyze your pages and pages of comments before lunchtime, but believe us—it is.

1. Start by downloading your survey responses from your online tool to Excel. (If you're still using paper surveys, you'll have to first manually enter all your data in the spreadsheet.)

2. Add a column to the Excel spreadsheet and use the random function. This function will randomly generate numbers for each record; copy the column and paste special to preserve the values.

3. Sort the column of your random numbers and select the first 50 to 60 rows to perform your analysis. Using a random sample ensures your sample of 50 will be representative of the whole group.

4. Go through the first 10 to 15 responses and develop categories for the responses. If you're doing a customer satisfaction questionnaire or user satisfaction survey and asked why they would recommend a particular service, perhaps your categories would be "Good support," "Ease of use," "Time saving," "Integration issues" and "Cost."

5. Quickly go through your last 35 to 50 responses and assign categories. If you don't know which category it belongs to, skip it and keep going!

6. Finally, sort by category and quality check your work. Each group should be able to stand together. As you read the groups again, you may find you can combine categories or create new categories. You'll usually be able to place the ones you skipped into a category.

Congratulations! You've preformed all the coding needed to proceed to make recommendations on the data.

And for those had to take the extra step of transferring data from paper to Excel, use the time we saved you with this trick to sign up for an online webinar to learn more about Cvent's online survey tools.

One area that many of our clients have questions about is reporting. As in, How do I make meaning out of all this survey data? What types of reports offer the appropriate insight in this case?

It’s no surprise that our relationship managers spend the majority of their day helping clients figure out ways to create meaningful reports. Understandably so, considering that the information contained within these reports is likely the initial motivation for their decision to conduct a survey.

There is no one right way to go about analyzing survey results, but there are a few best practices on how to analyze survey data that can really simplify the entire reporting process.

1. Work Backwards—Plan the Report Before Planning the Questions

We often tell clients to work backwards when planning their survey program. Rather than jumping into question creation, a survey planner should first have an idea of what the final report will offer in terms of useful information. If you are conducting a market demographics research study, and you are interested in knowing what age group is most receptive to an ad message, it is important that your survey ask age (or age range), rather than their date of birth. In the latter case, the same data can be extrapolated, but it creates extra work for you when it comes time for reporting.

Also, if you plan to include statistical information in your report (mean age, mean income, etc.), you want to make sure that your survey questions will elicit numerical responses. Don’t ask a respondent for their annual income with a one line comment box—your data will become very messy with responses like, "$35953.00" and "60000 USD." Again, know what data you want, and what you want to do with it first. Then, ask the questions to fetch this data according to your plans.

2. Use Filters to Drill Down—Segment your Data to Reveal Insight

Segmenting your respondents into targeted groups is a best practice we advocate often. It is just as important, however, to also segment the data when running analyses in order to reveal a deeper level of insight. For instance, your survey results may state that the average satisfaction rating from your entire client base is 4.7 out of 5. Upon further research, however, you might find that the average rating in most states is 4.9 or 5.0, except for the Northwest states, which gave an average rating of 4.1—clearly under company standards.

Without using filters to segment your data, you can overlook vital information that can affect your business. Filters can be used to separate and compare groups of data based on virtually any data point you have already in your database, or those that you might collect in the same survey.

3. Every Good Report Deserves Another—Observe Trends Over Time

No survey is an island. It might sound like extra work, but if you are looking to achieve real business results through your survey initiative, then your plan needs to include conducting the same survey again at a later time. Only by doing so can you observe trends and keep your organization accountable for the metrics you are aiming to manage.

If you are surveying your customers about telephone technical support, you want to ensure that you can pull cross-survey reports that show any progress or decline in feedback ratings. Such information is invaluable in directly managing a team; when such a survey is integrated into a team’s incentive plan, for example, these metrics can be instrumental in motivating employees to work harder and better.

Ongoing surveys that can be analyzed against each other with these cross-survey reports help you track changing opinions and needs so your organization can react efficiently to unexpected downturns, or perhaps increase revenue through new opportunities. Survey data, unfortunately, does have an expiration date, and in a dynamic business world, it is necessary to conduct ongoing surveys and accordingly adjust.