Is it time for M4D to get over the SMS hype?
Posted on November 25, 2015
Was that title a bit click-baity? Shame on me.
At another interesting meeting, this time at The Bistro, over butternut squash/watermelon/avocado salad, our conversation fell upon the failure of SMS projects in the governance and accountability sector in Uganda. Later, I began to wonder if it is, in fact, time to quit the SMS programs given the challenges, failures and lessons learned in M4D over the past fews. One of the issues I struggled with at a previous position was understanding the impact of SMS programs, if any, on the communities we were aiming to reach. BUT then I remembered my mantra:
CONTEXT IS KING.
CONTENT IS QUEEN.
While there are many drawbacks to using SMS as a communication tool, the benefits still outweigh these challenges and I do not think it’s time yet to abandon all hope for text-based programs.
What might the naysayers say?
1. Literacy and Language: Apparently ¼ of Uganda cannot read. 2015 Literacy rates for sub-Saharan Africa were reported at 64% and in South/West Asia at 70%. Women represent ⅔ of all illiterate adults globally. While I’m at it: only 1 in 4 of women in Central African Republic and Afghanistan can read! Furthermore, several languages, especially in sub-Saharan Africa tend to be more colloquial and don’t translate as well into text [Can SMS be sent in Amharic?]. Lastly, as mentioned in a previous post, countries with several languages can be quite challenging as it’s difficult to target SMS messages in different languages.
2. Content: 160-characters is really too few to transmit a message, especially on something like behavior change.
“Use condoms. Wash your hands. Don’t beat your wife. SMS *196# to Opt Out”
3. Shortcodes, reverse-billing and dealing with MNOs: Oh, the trauma. For users to ACTUALLY reply to you, oftentimes toll-free shortcodes have to be used. A short code may be specific to one mobile operator or “common” and supported by all major mobile operators. However, this then involves working on connecting the shortcode to each MNO, figuring out a reverse-billing process, working with the shortcode regulatory authorities etc., or else using an aggregation service. Depending on the country, it could be a breeze, or a soul crushing 2-3 year long process.
4. Spam: So.Much.Of.It. Want Love Quotes? Soccer SMS? Word of the Day? Unsolicited SMS messages has become such a nuisance that the Uganda Communication Commission made it mandatory for every single SMS of the sort to have an Opt-Out feature. People are more likely to ignore unsolicited SMS because of the misuse by various VAS vendors.
5. Low engagement rates: Getting responses on SMS surveys can be arduous, and this usually varies from country to country. In East Africa, people are very much tuned into SMS, however in West Africa, I’ve found SMS to deliver low survey completion rates. Adding in keyword responses further lowers response rates and introduces errors.
And the believers:
1. Cost: SMS are usually cheap-ish. Though, one might argue, that 30 seconds of a voice message could impart the same information as 3-5 SMS.
2. Reporting: SMS are great as a reporting tool. Thinking about the case of Rwanda where the Ministry of Health utilizes RapidSMS to get reports on maternal health + first 1000 days of a baby from community health workers. Health workers use a set of pre-determined codes to send quick reports back which are logged into a larger database. Simple and efficient, probably takes a few seconds once someone gets the hang of it. This similarly applies to any kind of registration or reporting in other sectors as well such as voter registration, stock tracking etc. No need for smartphones, computers, tablets, data etc.
3. Communicating numbers: It’s much easier to send an SMS of someone’s mobile money balance, than a voice message with the same information. As I learned in Luganda, 5-0-0 is apparently not 500. SMS is a great tool for delivering short information such as market prices, weather information, funds balance etc.
4. Useful for Directories: SMS can be great for directory-esque programs. For example, someone could use the first three letters of a town/district as a keyword to pull a set of information via SMS such as health clinics, retailers, polling centers etc.
5. Useful in low network and electricity areas: SMS can still be delivered to a phone within a time range once a phone picks up a signal or gets re-charged. This means messages can still be delivered when an end user’s phone is switched off. Most countries that I’ve traveled to in Africa unfortunately do not have voicemail features.
6. Storage of information on phones for reference: SMS information remain on phones, unlike voice calls which might be a bit more difficult to refer back to. For example, if someone received a text on the nearest health clinic, they could easily access that information multiple times by looking in their inbox.
7. Non-invasive: SMS surveys can be taken leisurely, at one’s pace. Rather than in one session over an IVR/USSD survey or an in-person survey.
In a nutshell, SMS is great within the right context, when using the right type of content, i.e.
-small bits of useful information that can either be reported or be referenced at a later time by an end-user
(reminders, prices, weather, directories, bank balances, electricity usage)
-programs where endusers are in control of what type of content and when they receive it (push vs pull)
It’s not a good idea for:
-lengthy surveys, soliciting open-ended questions/feedback etc.
-inclusion, given some of the literacy and language barriers.