Please don't annoy me! Five best practices for web intercept surveys and invites

By Jennifer Okula: VP, Data Innovation of Safecount

Web intercept surveys can be used for a number of different reasons including site satisfaction surveys, advertising effectiveness studies, or demographic profiling. Generally speaking, survey invitations can be displayed to site visitors while they are visiting a website, when they are leaving a specific webpage, or leaving a site altogether. The invite itself can take different forms such as pop-ups, pop-unders, full-page units, or DHTML floaters/sliders/overlays.
Web surfers do not want to have constant interruptions during their site visitation. However, they are also very willing to voice their opinions and value the ability to access content online for free. Careful consideration for user experience should be taken by website owners as well as research departments/companies. Here are my five best practices for administering web intercept surveys and invites:

1) Use a survey invite that is noticeable but appropriate
Choosing the right invite format is important. Audience and site content considerations should be taken into account. For example, younger or more technical audiences may use pop-up blockers so other methods may be more effective. Invites like overlay units should launch at the appropriate times. In the case of ad effectiveness research, they should be set to launch after, not while the advertising is showing. Invites should also be set to launch in the appropriate places. They should avoid launching over important content. Launching invites over video content for instance, would not be a good user experience. In general, the more "intrusive" invites, like full-page units, get extremely high response rates but consider the trade-offs with your audience.

2) Set the frequency and caps for invites appropriately

A website should have control over how many people or percent of page/ad impressions they want to launch invites to/from. A site may want to only launch invites from 5% of page impressions. And on an individual user level, a good rule of thumb is one invite per user per day, which is effective but not irritating. It would also be best if a user receives only one invite per day in total across all surveys running on a website, not just across one research vendor or survey. Additionally, if a user has already completed a specific survey, they should never see an invite for that survey again, instead of screening them out upon entry.

3) Use relevant and fresh creative content in the invites
Just as creative strength in advertising is important, the same principles apply to survey invites. Use imagery that might be more appealing to your specific audience. Is your site for pet enthusiasts? Include images of pets! Use simple and direct copy. Make it clear that opinions are wanted or feedback is needed. If using third party research vendors, co-brand the invites with the website name so visitors know the survey is approved by that site. Invite creative should be updated and refreshed often so repeat site visitors do not get accustomed to ignoring invites to new surveys that might be running.

4) Give consumers the opportunity to control their own experience or opt-out
Site visitors should be able to easily close a survey invitation if they do not wish to take a survey. The close or continue button should be prominent and easy to find. They should also have the option to choose how many survey invitations they receive or opt-out completely from surveys from research vendors.

5) Keep surveys short and engaging!
Finally, the best survey invites and methods can garner the highest click through rates, but if the surveys themselves are too long, you won't get the completes desired. In a web intercept environment, you are by nature interrupting a browsing or other web experience. Surveys should take around 5 minutes to complete in this environment. Keep survey instructions and survey flow very clear and as engaging as possible. Refresh the survey look and feel as often as possible as well.
You can find examples of much of the above in practice and additional solutions for helping manage web intercept surveys at my company's website,

» See the original article at