Text messaging - why and when to use text messages to change behaviour

26 February 2016Kim Louw

Categorybehaviour change

Tagsbehaviour change, text

Text messages have been shown to affect people’s behaviour.

Text messages can change behaviour

“But don’t you just send text messages?”

If you work in the behavioural insights field, you get used to this question. While the answer is of course, “no”, it is not an unfounded question as we do use text messages quite often. What can a lowly 160 characters do to change long ingrained behaviours? It turns out, quite a lot. In his recent book, “The 160-character solution: How text messages and other behavioural strategies can improve education”, University of Virginia Assistant Professor Ben Castleman shows that the humble text message can improve the educational outcomes for economically disadvantaged students.

Indeed our own work has shown that text message reminders to attend a hospital appointment increases attendance. This increased attendance led to a saving of approximately $70,000 per year for the hospital by using the most effective message. The savings could be even higher if they segment the population and use multiple targeted messages.

Although text messages have proven to be an important tool to nudge behaviour, it is not necessarily the best tool for every situation. So when should you text?

Text messaging: Why and when to use text messages to change behaviour

When considering an intervention, the mode of delivery can be as important as the message. Face-to-face meetings, phone calls, written letters, text messages, emails, and app-based notifications are examples of different delivery options to be considered when designing an intervention. Each of these modes has pros and cons so the mode should be tailored to the behavioural change the intervention is targeting.

Using the EAST framework designed by the Behavioural Insights Team in the UK, we map some of the pros of using text messages. 

  Public Policy Makers
  • Text messages are limited in length, making them short and sharp, and easy to understand
  • Mobile numbers change less frequently than addresses for some poeple - no need to update details to continue to receive service
  • Information is easily portable and can be taken to different venues, such as online tickets
  • Small time investment compared to phone calls and letters
  •  Text messages are cost effective (approximatley 15 cents per text)
  • Government infrastructure to send text is often already available (less upfront investment)
  • Text messages work on all mobile phone devices without modification
  • Mobile phone access and usage in Australia is very high, and unlikely to fall
  • The highest level of mobile telephone usage among young adults, socioeconomically disadvantaged populations, less educated young adults and people who rent or frequently change addresses - demograpics often missed by other modes
  • Undelivered messages bounce back so it is clear when someone has not received the message
  • Text message can be tailored to the recipient's needs
  • Can receive instant feedback/rewards for behaviours
  • Smart phones allow messages to open useful appsm e.g. Maps to show the route to an appointment
  • Text messages can be personalised and tailored to the individual
  • Easy to forward, share with friends, family on social media
  • Messages can be sent in bulk, creating a community of recipients of an intervention
  • Asynchronous - people can read the message when it suits them

Can be read closely to a desired action
Can receive feedback/reward instantly


  • Asynchronous - policy makers can automate the message to send when it suits them/intervention (e.g. isn't restricted to office hours)
  • Can specify when texts are received to prompt people when they are the most receptive
  • Can be sent when individuals arrive at key locations (e.g. a supermarket, children's school) or at specific times (e.g. at meal times, bedtime) 

Text messages have been shown to affect people’s behaviour in a variety of different domains, including payment of debt, smoking cessation, diabetes self-management, medical check-ups, and parental engagement with children’s education. These examples demonstrate where text messages effectively changed behaviours. But there are a number of contextual, situational and environmental factors that determine whether texts are the most appropriate mode of delivery for an intervention. A number of factors that make text messages a good option or poor option are summarised below.

Use text messages when... Reconsider text messages when...
  • There is a limited budget for the intervention
  • Recipients need to be reminded to do something
  • Information is short and sharp
  • People need to be given the information at a certain date and time
  • Recipients are from a highly mobile population that move frequently
  • Targeting a diverse population
  • When the intervention targets behaviour that requires many decisions and commitments over a long period of time, such as healthy eating 
  • When one simple action is needed, such as making an appointment
  • Information is confidential or sensitive
  • Legal information must be provided
  • Trying to convince someone of something controversial
  • Information contained in the message should be authoritative, such as first correspondence  of a fine or criminal offence)
  • There may be less mobile phone usage or access, such as remote Australia)
  • There may be an issue (or perceived issue) of an intrusion on privacy
  • The quality or reliability of phone numbers is poor

So when you see the next text pop up on your phone, remember that some serious thought may have gone into it!