Wiza Jakalasi on how African businesses can optimize USSD

USSD is one of the most popular channels available to people in Africa, but are businesses utilising it to its fullest? To learn more about this, we spoke to Wiza Jalakasi, current VP of Global Developer Relations at Chippercash, who shared some of his thoughts on how businesses in Africa can maximise this technology.

What was your first experience with USSD both as a technology and your personal experience?

Personally, it was on prepaid and airtime top-ups in Malawi. When the telcos introduced it, it was fascinating but I didn’t understand how it worked from the backend until I started working at Africa's Talking in 2016. We had USSD APIs, so I got to understand what was happening at the granular level, step-by-step. Then we started building apps on the USSD API and encouraged others to build similar solutions in Malawi. That’s where I first got into it.

Three years later, I joined  Hover where we got into exploring USSD as a data transit layer, and I learned a lot more from that. I would say I’m pretty well-versed with all the different things that can happen on the stack.

Tell us a bit more about the USSD APIs at Africa’s talking.

We would help businesses generate static USSD codes, but they could modify the logic by adding subcodes to make the codes do more.

You mentioned that while at Africa's Talking, you encouraged businesses to build with USSD. What kind of products did they build?

Most of the apps were very simple apps for stuff utilities. Utilities to query for balance or account status, examination results, or any sort of information that could be expressed in the form of an SMS message, etc.

For Malawi specifically, there weren't a lot of online alternatives to get that information because internet penetration was (and still is) extremely low but mobile penetration is relatively high.

So, dissemination of weather information, health information, and basic services through USSD were areas where we saw a lot of activity and growth. We did a massive national campaign with the Ministry of Health with regards to tracking the spread of non-communicable diseases.

The use of USSD was pretty much about the dissemination of information and short services across different sectors that I have described. It made sense for our region because not a lot of people were online. It was not viable for businesses to pursue online products because people would not be able to access them.

When you were trying to get businesses to use USSD, were there any pushbacks?
Yes. Some businesses didn’t understand the value or they didn’t see how their services would fit into the context of USSD. There was a little bit of education that needed to take place there.

Another pushback was that it is expensive to develop and maintain the system because they had only dealt with telco and telco engineers. And telcos would charge them a lot more because they ran an enterprise model.

The advantage with APIs like Africa's talking was that they made it possible for people to use their web development skills to build USSD apps. That paradigm shift changed when the USSD API came into play.

Just how expensive was/is it for businesses to set up a USSD channel?
I think in the absence of API providers and aggregators, it is pretty expensive because the telco rates are too high. However,  things have changed in the past five years. Today, there are aggregators in almost every African country providing USSD Codes.

Now, I don’t think cost is the issue. For certain businesses, you can argue that it is even more expensive not to adopt USSD considering the addressable market size that you’d be ignoring. Businesses cannot claim that cost is the issue anymore.

Are there certain businesses that are more likely to adopt USSD than others?
Yes, any business that has thousands of customers, it makes a lot of sense for them to use USSD. For example, banking, insurance, healthcare, utilities, and security.

For any business that needs to coordinate at scale, USSD is still the best to get to a large-scale population on the continent.

If you look at internet penetration, there’s a big gap between availability and use. I think there are about 400 million people across the continent who have access but aren’t using the internet. Those are people you can reach via USSD.

Is there a role for USSD in small businesses? Is there a possibility for adoption? Does it make sense for them to even adopts USSD?

I think it makes sense at a specific price point but not at the price point that is general to every type of business. For example, a big restaurant can use a mobile app for reservations since most of the customers are mostly smartphone users so you can probably design a more elegant experience on top of the mobile web

But if you’re a small restaurant owner whose users don’t own smartphones, you may think USSD is useful but there is no need for it because of the cost.  I think USSD is really for larger businesses trying to reach people on a large scale. I don’t think there are use cases for smaller businesses.

What would a fully-optimized USSD experience look like for businesses?
I think there's a specific context that it performs extremely well like parking.

When you are paying for parking tickets in Kenya most places have a USSD code that you can dial.  It is integrated with other internet connecting systems where you enter your vehicle plate number. The plate number is already in the database of the parking provider which was captured by the camera by the time you were entering the parking lot and matched with your ticket. So, when it’s time to pay, you dial the code and enter your vehicle plate number which was already captured. When you press the pay option to make payment for the ticket, it ends the USSD session and invokes an SDK push through Safaricom’s MPesa express API and you input your pin to conclude the transaction.

It is a mind-blowing experience and only possible with USSD because it supports a wide range of phones and it combines the offline performance of USSD with the convenience of the internet in a very seamless way for the users.

I think that’s a key example of what USSD looks like when it is on steroids.

Do you think the businesses that have adopted USSD are optimizing it?
No, not yet. USSD is still unreliable in some cases, in other cases, it’s not an elegant experience.

Another example of an optimized USSD experience is in South Africa where banks send you a USSD push with the details of a transaction, and give you the option to either authorize it or decline it. That’s two key presses instead of an OTP that requires up to 8 key presses, especially if the pin is 6 digits. The process is definitely faster than using an OTP.

There is a lot of room for further optimization when people fully understand how USSD can be combined with other online technologies to deliver novel experiences.

You mentioned the reliability of USSD as one key thing to improve, how can businesses improve reliability?
I think the big thing is reliability issues which can be either on the side of the provider or due to infrastructure issues at the backend of the businesses. For example, if you’re looking up something in the database, on the web, the user can wait for it to load. But on USSD, if you can’t return that result within a specific time frame, the session will time out, and that can be very frustrating for the user.

That said, reliability is mostly contingent on the providers, and in some cases, they don’t even have that level of visibility because they’re reliant on telcos. However, what businesses can do is optimize the response time for the USSD codes as much as possible.