The most important part of a web survey is arguably the email invitation. The invitation sets the stage for the rest of the survey, and allows the respondent to judge the costs and benefits of participation. The content of the invitation can also determine if the invitation gets delivered or blocked by email providers, and whether it lands in the recipient’s inbox or spam folder. As web surveys increase in popularity but continue to have the lowest response rates of all modes, it is important to understand how the features of the email invitation impact non-response. Many features of the web invitation have been tested in past studies, such as subject line, personalization, contact and confidentiality information, topic, sponsorship, frequency of contact, incentives, and URL links. However, results have been mixed, and many studies were conducted in academic settings with unknown applicability to general population settings. Further, many studies have tested one feature versus another, but have not tested the combination of multiple features of the email invitation. The present study expands on the current body of knowledge by simultaneously testing multiple features of the survey invitation. These features include subject line, incentive, HTML versus plain text, sponsorship, personalization, and frequency of contact. The invitations were sent for two different types of general population studies. The first is the Gallup Panel, a probability- based panel of 55,000 members who receive survey invitations on a regular basis. The other is a customer satisfaction survey, in which respondents have a customer relationship with a business but are not expecting a survey invitation. The findings will include results on bounce backs, SPAM flags and spamminess scores, open rates, click through rates, and final response rates. Recommendations are also made for features that improve deliverability, open-rates, click through rates, and ultimately response rates.