Six Variables of Online Surveys

Previously, I looked at the need to combine detailed data from site analysis systems with additional consumer insight gleaned from surveys. Today, some thoughts on how to ensure your surveys are as effective as possible.

There are many different forms of surveys that online businesses might run, and they can vary on a number of details. You might survey site visitors as opposed to customers; you might collect general background information; or you might ask about a specific issue. The survey might be a one-off or run continually. Some online survey variables include:

  • Purpose

  • Target audience
  • Survey type and how respondents are recruited
  • Number of responses needed
  • Expected response rate
  • Survey frequency

Survey Purpose

A survey should have clear objectives; there must be a reason why you want to do it. There could be more than one research objective in a survey, but they must be clearly stated, easily understood, and consistent. From the objective flows everything else: the type of survey needed, the target audience, and so on.

The Target Audience

The objective should make it clear who you want to survey. You probably don’t want to invest time and effort understanding everything about everybody who visits your site. Your primary interest will be about learning the right information about the visitor types that are of the most interest to you, such as customers and subscribers.

Survey Type

Having determined the survey’s objectives and audience, you can better choose the type of survey that will meet your needs.

There are two main types of online surveys: pop-ups and site-based.

A pop-up survey (as the name implies) pops up in a window on your site. Generally, it must be short and easy to answer. A site-based survey is potentially more extensive and people are directed to a separate part of the site or another site altogether. Here are some key differences between the two types:


Pop-Up Survey Site-Based Survey
Pop up on the site Hosted elsewhere on the site or on another site
Short (about 5 minutes to complete), as they are invasive Longer (up to 15-20 minutes)
Susceptible to pop-up blockers Not susceptible to pop-up blockers
Survey invitation is generally random on the site. Survey invitation can be specific (by email) or random (on the site).
No control over who answers the survey Control over the number or type of people who answer the survey

Number of Responses

Another key consideration is the number of completed responses you need. This can vary enormously with your type of work and the survey’s target audience. For consumer analysis, you’d ideally need about 400 respondents to do any meaningful analysis.

Response Rate

Having determined how many respondents you need, think about how you will get them. For a pop-up survey, visitors are typically randomly selected on the site and presented with the pop-up survey invitation. For a site-based survey, people will be invited by either email or an invitation on the site.

In either case, only a portion of those invited will actually complete the survey. This is known as the response rate. The response rate can vary from survey to survey and is influenced by:

  • The style and quality of the survey’s first page

  • Potential respondent’s relationship with the site, brand, or both
  • Interest level of and survey relevance to the potential respondent

If you use email to invite people to participate, the email’s style and subject line will also affect the response rate. Make the call to action as interesting and engaging as possible, so it cuts through inbox noise. Use language appropriate to your business type and your relationship with the potential respondent. Many email and survey systems allow you to personalize the invitation. Use this to improve chances someone will open the email and act on it.

Survey Frequency

The final consideration is how often you will run the survey. Many surveys are run only once to get some insight into a particular issue, such as the effects of a new site design. But some surveys, such as a customer satisfaction monitor, run more than once or continually.

And there you have it. Happy surveying!

Related reading

tencent_emily-ma_featured-image
kenneth_ning_emarsys_featured-image
bounce-370x229
site search hp
<