Best Practices When Using Text Messaging in Survey Research

For over two decades, online survey panels have served as an option for researchers to reach survey respondents at a generally lower cost when compared to the cost of live telephone interviews.  Put simply, an online panel is a group of internet users who have been pre-recruited and have agreed to participate in numerous future online surveys.  Companies who “sell” online panels make the investment to recruit respondents, and then charge researchers for access to the respondents.  Because the recruiting costs can be spread over multiple projects, the panel cost for any given research to “access” a respondent is relatively low.  The more frequently a panel company can sell access to respondents, the more profitable it is for them.

The Panel Problem

A huge problem for online panels is that they are a very finite resource.  When the desired group of survey respondents is distributed in a geographic area smaller than an entire state, online panels become much more limited in the number of surveys they can produce because of limited number of panelists within any given state.  This limitation becomes exponentially worse when geographic boundaries shrink to a regional, county or city level.

Evolving Beyond the Panel

Several years ago, American Directions Research Group recognized this limitation of online panels and began to explore alternative means of reaching respondents and persuading them to complete an online survey. 

Over 85% of the U.S. population uses a smart phone.  There are 294.15 million smartphone users in the United States.  Approximately 47% of web traffic in the U.S. originated from mobile devices.

Cell phones are our constant companions. As many as 71% of Americans check our phones within 10 minutes of waking up. And 74% of Americans cannot leave cell phones at home without feeling uneasy. 

Most importantly for our purposes, 70% of Americans check their phones within five minutes of receiving a notification.

The Process

• Beginning with a calling list drawn from either general population or voter registration lists, phone numbers are checked against a constantly updated database of known cell numbers.

• An online survey is programmed and hosted on ADRG servers.

• A manually-initiated peer-to-peer text is prepared for each respondent, containing a unique URL (web link) to the online survey, and these are then sent to respondents.

• Upon receiving the text, a respondent simply clicks on the unique link and is immediately delivered to the online survey.

How Different Demographic Groups Respond to Text Invitations to Surveys

American Directions has manually initiated millions of peer-to-peer (P2P) text invitations to research surveys over the last several years.  In the course of these efforts, we have amassed a body of knowledge about how different types of respondents react when texted for these purposes. We have developed various best practices to maximize the number of respondents who participate in surveys.

Survey response to text invitations can vary amongst different demographic groups.  Sometimes, the response to text invitations for a given demographic can also be different from that group’s response to a live survey call.  This type of knowledge, coupled with a true multimodal survey platform such as the ADRG Voxco system, allows clients to select the most appropriate channel for any specific demographic group.  Here are some examples of what our processes have revealed:

• Gender: Men are 11% more likely than Women to respond to a text invitation, but only 2% more likely to respond to a phone survey

• Age: While newer technologies such as smartphones are often associated with younger users, out of all age groups, older users are the most likely to respond to a texted survey invitation.  Respondents aged 65+ are almost twice as likely to respond in comparison to 18-34 year olds.  The 35-64 age group is about 40% more likely to respond than an 18-34 year old.

• Ethnicity: Whites are the most likely to respond to texted survey invitations, two times more likely than Asians, Hispanics or Native Americans, and approximately three times more likely to respond that African Americans.  In this case, a live telephone approach would be more appropriate to reach African-American respondents.

• Party Affiliation:  Our data shows that Democrats and Republicans are equally likely to respond to a texted survey invitation.

• Education Levels: Texting response is generally consistent across all education levels with the exception of those who have not graduated high school.  Those respondents are about half as likely to respond to text invitations than respondents with all other levels of education.  When targeting lower education levels, it is more effective to use live telephone interviewing versus a text invitation approach.

Making the Most Out of Every Invitation

Best practices such as these which have demonstrably improved survey response rate include the following:

Personalization: Whenever possible, ADRG includes the name of the potential respondent in every invitation.  Personalization keeps a human touch with respondents. Another example of how to lift participation is to prioritize the use of “I” instead of “We” to emphasize communication from one human being to another.

Getting to the point: The attention span of potential respondents is very short. It is important to write a clear and concise message to prevent to avoid losing the respondent. In a few lines, the recipient should know the subject of the SMS, as well as the purpose of the survey.

Association with Organization or Topic: Make clear to potential respondents what the survey is about.  Letting them know the survey is about current issues in their local school district helps respondents relate better.  When surveying members of an organization such as a Union, this sort of disclosure is even more critical…

“…Your union, the AFT, is conducting a short survey on important issues facing public school staff and faculty. Your individual responses will be kept entirely confidential. Please make sure the voices of school employees are heard, and help us advocate for you, by clicking on this link to take the survey!”

Include a Call to Action at the end of your SMS: Once your invitation to respond to a written questionnaire is completed, it is important to include a Call to Action. At the end of message, the URL of your survey is included, along with a sentence cordially inviting respondents to complete it. Space around the link is important, to allow the respondent to spot the CTA at first glance.

Use a local phone number : One of the first things your customer will see when they receive the SMS is the regional indicator. ADRG sends from area codes in the same region as the respondent. Sharing the same regional indicator builds respondent confidence that they can trust the SMS, which increases opening rate.