To measure user satisfaction with the 2020 Census online questionnaire, the U.S. Census Bureau re-contacted a sample of census respondents via text message and asked them to complete a short online survey. The representative nationwide sample was drawn from online responses to the actual census where a valid cell phone number was provided. Each sampled phone number received up to 3 text messages throughout an 11-day field period. We embedded a fully crossed experiment within the sample to determine whether the time of day the text was sent (12 noon or 6 pm) influenced overall response rate or satisfaction. Approximately 5 percent of the sample responded to the survey and 15 percent opted-out of text messaging. Most responders reported the day they received the text with a drop the following day. The response to the second text was similar to the response to the first text, with only the third text generating a lower rate of response. Finally, there was evidence of more attentiveness to the noon condition versus the 6 pm condition as measured by more responses but also more opt-out requests to the texts. This paper shares these findings and other lessons learned from a text-only survey notification contact method.