Skip navigation
NEWSLETTERS  |  SEPTEMBER, 2002

Online Surveys

By Eric Holter

Online Surveys



A manufacturing client wanted to know if their customers would be interested in buying a limited set of their products online. It was going to involve a sizable investment to create an e-commerce site and, in addition to the costs, there was the risk of upsetting some of their retailers and distributors. Was it worth the risk and expense? We talked it through a few times and solicited input from various departments. While everyone agreed that it seemed like it would be well worth the cost and effort, the real opinions that mattered were those of the consumers. Would they buy online? To find out what they thought, we created a survey. The survey contained roughly eighteen questions designed to determine the customer's willingness and comfort with purchasing our client's product online. Our client had used surveys in the past, but found them limited in their effectiveness. They required considerable costs to mail out, and, in addition to the costs, the completed surveys took a long time to come back. When they finally did come back, the percentage of respondents was often not very high. The task of compiling results required a lot of work. Someone had to do data entry to get the written results into whatever software was being used for analysis.

We thought, "since we're trying to determine the comfort level of their customers in buying online, why not use an online survey instead of the traditional survey?" We spent some time building the survey forms and some simple reporting tools for results. We set the survey up to email the client with the survey results as surveys were completed. The client had the email addresses of many of their customers, which they had collected from a monthly email newsletter offered on their website. They decided to offer respondents a chance to win some of their products in return for spending ten minutes filling out the survey. We sent out the email to about 8,000 recipients. Then an amazing thing happened. Within five minutes of sending out the emails, results started coming in. Out of curiosity, I took a look at two or three of the initial emails. As I was reading them, three more came in, then five more. Within a half an hour we had over two hundred survey results. By the next morning we had over two thousand! We ended up with at least 30% of the recipients responding to the survey. Aside from causing the client's email box to overload, the information provided the client with much better data from which they could make a more informed decision about selling their product online.

Another example

Another client approached us about creating an online survey. Their survey was much more complex than the one we had previously built. It required using various types of survey question formats. Among them were multiple-choice questions, choose many or choose only one type questions, interval questions (1 being bad, 10 being great), and open text response questions. Our client wanted to send an email to recipients offering them a chance to win a product for filling out the survey. In this case they did not have an internal source for the email list. Instead they purchased a few different lists from highly targeted opt-in email list brokers. Similar to the first survey, this one generated an amazingly high respondent to recipient ration (over 20%). The results started pouring in within minutes of the email going out. At many points throughout the day that the survey when out, there were over 100 simultaneous surveys being filled out! This data was extremely valuable to the client in making marketing decisions.

Adding Online Surveys to a Website

Building a NewfangledCMS application

As with most of our custom programming, we decided to invest additional time of our own in building a survey application for our client so that we would have a NewfangledCMS application to offer our existing and future clients. The result was a highly robust survey building system that allowed our client to build, modify and distribute survey with minimal input from us. We are now able to offer this component to other clients. It requires a one-time installation and then provides the ongoing capability of building an unlimited number of online surveys.

How the Survey Manager works

We built the Survey Manager so that our clients can create online surveys without any help from us. While creating surveys is extremely simple with this tool, we did add the functionality for doing some advanced survey logic (such as "pathing," directing a respondent to specific places in the survey based on their answers to various questions).

While writing a good survey does take time and thought, putting the survey online takes only a few minutes with the Survey Manager. Here's how it works. As with all NewfangledCMS applications the client first logs into the system. Under the "Do Stuff" menu, the client then selects "add new survey." Having named the survey, the client now adds questions through the Survey Manager interface. This interface uses the familiar and intuitive NewfangledCMS controls. The first step is to add the introduction text and conclusion text for the survey. As with all NewfangledCMS content this text can be easily modified as needed throughout the process.

You can add questions to the survey by selecting "Add question" and choosing from a list of eight types of questions. These eight types cover the bases for creating most survey questions. However in the event that a different question format is required, we can easily create custom question types on a case-by-case basis. The eight types include multiple choice, multiple choice with images, open text field, multiple text fields, intervals (1 being bad, 10 being good), intervals with images, multiple scale intervals, and pull-downs. The multiple choice formats allow you to choose "select only one" by utilizing radio buttons, or "select as many as apply" by using checkboxes. These question types also have an option for an "other" field that creates a text field where the respondent can specify what their own response.

The questions are entered through the standard NewfangledCMS content entry screens. The various fields for creating the questions are all documented and include help windows to remind you of the proper format for entering information, and what the various options are. Using these tools you can add as many questions to the survey as you need. Of course the number of questions, while technically unlimited, should be appropriate to the audience you are sending the survey to, and the motivation (premium) you are offering to them for filling the survey out.

Once the survey has been entered, the system provides additional survey management tools. One of these tools allows you to copy entire surveys. These copies can be modified to create alternate versions of the survey with slightly adjusted questions for different audiences or as benchmark surveys for better analysis of the primary data collected from the main survey. The survey questions themselves offer optional benchmark questions and benchmark frequency. Another survey management tool is the ability to append unique identifiers (variables) to the survey's link. This identifier will show up a in the results data to allow you to distinguish between various groups of respondents. The survey manager will also remember where a particular respondent is within the survey, so if they aren't able to complete the survey in a single sitting they can return at a later time to answer the remaining questions.

While the survey is running you can "view stats" which helps to monitor survey activity. Stats lists the total number of survey sessions, the number of completed surveys, the total answered questions, and the number of current active users.

All survey results are stored in the NewfangledCMS database. You can download the results as frequently as you want. They are downloaded in an excel spreadsheet in standard column format with each question along the top and a row for each respondent's answers. This data can then be imported into whatever tool you may use for data analysis.

Hosting

The "Max Number of Users Allowed" throttle is used to level off server usage when clients send out a survey. There is often immediate heavy traffic to the survey as respondents fill out the survey. The traffic typically dies down over the course of a day. Depending on how many people receive the email this traffic can overload the server. By adjusting the number of simultaneous users we can choke off overload and protect those survey users alreay in progress. The "Max Number of Users Allowed" can be adjusted based on how many people the survey is being sent to.

Conclusion

Newfangled is committed to helping our clients to use the web as an effective business and communications tool. Online surveys can be a powerful tool for gathering customer feedback, doing product research, or refining other aspects of your business's direction. Using the NewfangledCMS Survey Manager makes the creation of surveys quick and simple, which provides one more way for us to help you and your clients maximize the value of the web.


Comments
Christian Rathbone | March 22, 2011 7:24 AM

I myself didn't want to take up surveys. They ate up my time and definitely never been in my priorities to consider taking one. I could only take up surveys if they guarantee something in return in exchange for the time spent.