With the overwhelming optimism around mobile phone-based communication technologies, it is reasonable to ponder: Do literacy levels pose a serious limitation to SMS-based communication campaigns?
Imagine the following scenario
It is a busy Monday morning. You’re rushing to get breakfast on the table, to pack lunches for your kids, and to get everyone — including yourself — out the door. Your phone pings:
You skim the SMS, and think to yourself…
I don’t know why I’m getting a message in Spanish, I only read and speak English… Perhaps this afternoon I can ask my Spanish-speaking neighbor what this means?
Shoot, I see the word Zika: Something something Zika something something something something. Ugh, I can’t deal with this right now, I need to get the kids to school… We are running late.
If you were able to read “Zika” plus one or two other words in the scenario above – congratulations! Having read part of that simple sentence, you are now considered “literate” in Spanish, according to Demographic and Health Surveys (DHS) criteria.
The DHS classifies “percent literate” as the percentage survey respondents who attended post-primary school or higher, and those who can read all of a simple sentence, and those who can read part of a simple sentence in local national language.
In Zambia, where I work, it is with this formulation that DHS arrives at 68 percent literacy.
Literacy Statistics in Action
Recently I spent the day in a crowded Lusaka compound interviewing three teenage mothers.
Outside the tiny one-roomed house, the maze of dirt roads hummed with activity, kids ran barefoot, barely dodging streams of open sewage, dragging behind them kites made from tattered plastic bags.
Inside, the three ladies were shown cards with simple sentences in Nyanja: “Crops are your livelihood,” “Lives of mothers and babies are precious,” “Call if you want to know how to take care of your baby and yourself.” In the dimly lit room, each of the women stared blankly at the cards.
None of these women could read. Not a single word. Not in their local language, Nyanja, let alone in English. Only three months earlier these same young mothers had received this message from MTN, a Zambian telecom:
How do you think the young mothers reacted? Simply, the ladies didn’t react; they couldn’t understand a single word.
The Explosion of SMS-based Communication Initiatives
As Zambia and Botswana Program Manager for Human Network International (HNI), I am clearly a proponent and practitioner of ICT4D. SMS and IVR projects are a big part of what we do.
With the ever-increasing mobile connectivity globally — particularly in some of the world’s poorest and least developed countries — it is no wonder that mobile phone-based communication initiatives have quickly become the hip new communication tool for international development agencies.
With SMS, is it now possible to quickly and efficiently disseminate public service messages on a scale simply unattainable by other communication platforms such as posters, radio spots, pamphlets, or billboards.
SMS Campaigns Overlap with Illiterate Populations
One appeal of SMS-based communication campaigns is their ability to reach massive swaths of the populations — particularly isolated, rural, women with limited education in sub-Saharan Africa and in South and West Asia. The exact audience that is least likely to be literate.
According to the UNESCO Institute for Statistics, the lowest national literacy rates are observed in sub-Saharan Africa and in South and West Asia. In 14 countries, less than half of the adult population is literate.
mHealth Campaigns: Adding to the Confusion
Many SMS communication campaigns are of the “mHealth” variety. Patients receive messages to remind them of appointments, and to increase health information and service access.
Low literacy poses a serious barrier to information comprehension, particularly in instances of health and/or scientific terminology. It should be obvious that sending health-related SMS’s with words such as ‘meningitis immunization,’ or ‘schistosomiasis’ to those who are illiterate is a futile effort.
However, mHealth SMS initiatives are designed to disseminate medically-focused messages, which are often outside the comprehension of even those men and women who can only read part of simple sentence. Again: These people are considered literate by the DHS.
SMS Campaigns vs. DHS Literacy Statistics
In Mozambique, Ark piloted an SMS system intended to improve HIV treatment adherence and prevention of mother-to-child transmission of HIV. However, in Mozambique, 68 percent of women either cannot read at all, or can only read part of a simple sentence.
In Mali, TCC Mobile supported the Ministry of Health to develop an SMS platform to prevent the spread of Ebola. But in Mali, 79.2 percent of women and 63 percent of men cannot read at all. An additional 4.6 percent and 8.4 percent of women and men, respectively, are able to read only part of a simple sentence.
In Ghana, The Mobile Midwife Application enables pregnant women, new mothers and their families to receive weekly SMS and/or voice messages. Nonetheless, in Ghana 32.8 percent of women cannot read at all, and an additional 1.9 percent of women can read part of a simple sentence. So, in Ghana, a country with relatively high overall literacy compared to many neighboring countries, a surprising 99 percent of enrollees chose to receive the messages by pre-recorded audio.
Effective Technology In Spite of Low Literacy
Back in the bustling Lusaka compound, each of the three young women — the ones who stared blankly at the written messages — were given the chance to beta test messages on HNI’s Interactive Voice Response system, the 3-2-1 Service. Listening through the speaker of the simplest mobile phone, the ladies’ eyes lit up as they heard a native Nyanja speaker giving tips about feeding a 6- to 9-month-old baby.
In a matter of an hour, I saw women go from being completely information-isolated in a world of written communication, to being empowered by an audio information system.
Experiencing this, I became highly skeptical about SMS communication campaigns. They are not the most effective mobile phone-based approach to reaching audiences that may be considered “literate” by an easily misunderstood measure, such as the DHS. We run the risk of prioritizing seemingly efficient campaigns that in reality have little impact beyond high numbers of “beneficiaries reached.”
As we know, technology doesn’t wait for anyone, and mobile phones are entering the hands of more people each and every day. While this will rapidly outpace what are traditionally slow gains in literacy, it still provides a unique opportunity to reach previously isolated populations with important information in a language they understand.
It’s time to think outside the SMS box, looking to other, more appropriate ways to reach populations, like the three young illiterate Zambian mothers, via mobile phones.
Olivia Bell is based in Lusaka, Zambia with Human Network International.
YES! Read our blog post about how IVR, not SMS, is the solution for Change: https://www.engagespark.com/blog/sms-really-solution-change-voice-calls-are-for-most-poor-people/
To learn more about when to use IVR vs SMS, check out: https://www.engagespark.com/blog/make-biggest-impact-voice-sms-text/
And this is exactly why we created engageSPARK; IVR is the most effective way to *interact* with constituents (especially those who are illiterate), but has traditionally been expensive (large setup and monthly fees) and requires an IT person or lots of training. We wanted to build the world’s easiest to use and most affordable solution: non-technical staff can actually build IVR applications in just minutes in any country.
FYI, we’re a not-for-profit social business committed to helping NGOs of any size leverage IVR technology to interact with anyone, anywhere with a phone. No subscription plans or implementation fees; low usage fees only.
Interesting. Despite these three women, I think you’d be hard-pressed to find a medium with the ROI of SMS information, which is close to free. Organizing and recording voice and then calling is not. Even if an SMS system isn’t reaching absolutely *everyone*, with limited resources, it has the best chance of reaching the most people. Not to say other methods aren’t needed, but writing off entirely a technology because it doesn’t accomplish all possible goals seems a bit extreme. My bet is that robo-calls are far more irritating and likely to be ignored after the initial novelty wears off, too.
Clarification: Not all IVR-based systems are robo-calls. The HNI 3-2-1 Service allows people to call the service and “pull” the information when they want and need it. In the event where an alert is necessary, a “push” is possible.
Re. the ROI of SMS: Yes, it’s cheap. it’s a cheap way to reach the people who least need the information — the thin demographic slice of people who are literate and likely have other ways to access the information. In some countries that slice is thinner than in others, and that should be a consideration when designing a project.
For me, the most telling anecdote is the Ghana The Mobile Midwife Application — where 99 percent opted to receive the messages as voice recordings.
I would love to see some behavior change research done to determine the actual impact of SMS-only projects vs. IVR-only projects vs. projects where the end user can choose a preference. If this research exists, I’d like to know about it. But that would be a better way of getting to the ROI question.
My hypothetical multi-country study:
If $10000 can reach Y people by SMS but only one-quarter or one-tenth as many by IVR, how much behavior change do you get in each scenario? (I just made those numbers up, BTW). My bet is that the behavior change efficacy of SMS would track strongly with the literacy of the target population. My other bet is that the efficacy of an IVR approach would have more behavior-change-per-dollar regardless of the literacy rate in any of the countries.
Good point about the ROI and the information delivery throughput;
It is a matter of deeply knowing your segment ( best shown by the test of cards towards the 3 women, obtained on the field) and by simply testing SMS/Voice interactive scenari a/b at low scale before any upscale.
This can be done by simply accessing to enough powerful tools for mobile interactivity with enough user friendly interface and easy pay as you go business models.
There are many ways indeed to find cheap SMS , generally leading to design information services which are so cheap to be of any value …. typically, this services do no support interactive (2way interaction) which prevents the service designer to make any test, and therefore to have any idea at all about what is the understanding of the message received, as there is no way for him to simply measure any Return, which could start at triggering back any Call to Action upon receing the incoming message.
There are therefore so many projects spending huge amount of money on lowvcost SMS services (typically MSS one way), which therefore leave one totally blind about what is the real efficiency of their campaign (not to talk about those who structurally are not interested, or nor committed, to investigate into Returns and Benefits on the field).
There is still plenty of room for dual approach in service logic, at reasonnable cost, combining SMS and VOICE (with multiple vernacular languages in voice for example) provided using flexible tools which combine in whatever needed way : SMS, Interactive SMS, outgoing voice, outgoing interactive voice, incoming voice, incoming voice & record & bridge to third party, within the same campaign tool .
This is typically possible from the Cloud without traditionnal CapEx , enabling Programm Managers to experiment and leverage progressively the level of interactivity offered.
It is cheaper to start with SMS broadcast ,but SMS interaction at least, or combined IVR interaction should be included from start in the service logic to build first step for any Return.
In doing so, you get :
-measure of interest or understanding from sendees
-measure of appetence for more information and further action
-additionnal (with more expenses, but justified) opportunities to deliver more accurate or dedicated message in test or voice mode
-possibilities to interact with the user, as well as record, monitor, redirect, ofer alternatives, advocate for further information available, etc.
Of course, this will also firstly give you a good idea about acceptance of your input (the original first message in SMS MT for example) …. but better know this the sooner, than possibly continue to spend money regularly on non efficient SMS campaigns !
Typically you might use in initial SMS message, basic words and sentence in wordings which are as universal as possible in whatever vernacular language and literacy level, and deliver more information upon SMS answer collected from destination party, also suggest inbound number to be called for more information, support outbound interactive call TRIGGERED upon receiving some SMS request (therefore this is not a robo call, although it is fully , but smartly , automated call, delivering appropriate info for your programm …and also pointing out phone numbers which are receptive to your message, possible good targets fur further more advanced info and leveraging mobile interactivity (other SMS campaign, pure outbound new voice campagn for deeper dissemination of information from your programme, etc etc).
However, the long time habit of using simple lowcost oneway SMS gateways, instead of more recent, flexible and versatile mobile platforms available simply from the cloud without CapEx , still prevent service designers to benefit from interactivity of duale SMS/Voice services, bringing higher penetration and dissemination of desired message, delivering valuable feedback and data on interest and acceptance, and …. bringing finally much more attractive ROI without difficultie for such more smarter programme deployment over mobile.
If interested , you may explore sites like http://www.CloudCaaStor.com where you will be able to start on a DIY basis outgoing SMS and Voice dual campaigns, for better ROI and deeper penetration of benefits to be disseminated by your programm.
On HNI’s blog we have a new post that is sort of an “Appendix” to this article, including a peek at the DHS data that led to a change in how we respond to requests for SMS projects.
How we became SMS Skeptics