E-Agriculture

Question 2: What are the barriers to reaching scale with mobile agriculture information services and...

stephane  boyera
stephane boyeraSBC4DFrance

Launching the discussion on the second question.

My experience with market information system as well as with helpline services is that the scalability comes from the architecture of the service far more than a relationship with  mnos.

Let me take 2 examples:

*services relying on intermediaries

the model on intermediary-based service is very popular, particularly in market information system, mostly because of the technology: sms is not usable by most of the farmers living in the developing world (if you are interested by number see e.g. a post i wrote few weeks ago: http://www.webfoundation.org/2011/10/review-of-the-new-vodafoneoxfamaccenture-report-on-mobile-for-agriculture/ ) and today the vast majority of such systems are sms-based. So the implementation requires the use of literate, sms-aware intermediaries. So then, one need to idenitfy potential intermediaries, train them, pay them as part of the model. This is a problem from a sustainability perspective, but also a problem from a scalability perspective: finding and training intermediaries at a country wide level. This is working perfectly at a local level, but I'm not aware of any of these systems that have succeeded to scale up.

In my opinion, the only option is to work towards removing the need for intermediaries in such system, and this is more a technology problem, at this point in time, rather then anything else.

*helpline service

Second example of service: helpline. here, no intermediaries for farmers (thus the success of such services on the user side), but on the service provider side, the human-handling process is killing the service. A knowledge worker is able to manage up to 70 requests a day, that pays as one can imagine a very small part of their salary. So here again, the problem is in the architecture of the service. Such helpline can only be sustainable and therefore scalable if knowledge workers are replaced by technology (see also the same post i mentioned above)

So as i pointed in another post in question 1, i believe that partnership with mnos is not a magic wand that would make a non-sustainable non-scalable service becomes very profitable.

steph

Judy Payne
Judy PayneUSAIDUnited States of America

 I am curious about Stephane's comment that SMS is not usable by most farmers in the developing world.  Certainly illiteracy is a challenge but I understood that even illiterate farmers found using SMS for market price information acceptable.  Or they were able to find a literate relative or friend to help them.  One solution is indeed intermediation as Grameen CKW does.  And the CKW model is working to hone a business model to make it sustainable and scalable (a key element of this is to use the CKW's -- community knowledge workers -- to collect information on behalf of paying clients).  

Certainly voice based services are more convenient and more and more services are finding a need to add this option but keeping call center costs low enough to be sustainable and scalable is a significant challenge.  Has this been done anywhere in Africa?  Or are the Indian examples (perhaps IKSL) transferrable?

Judy

stephane  boyera
stephane boyeraSBC4DFrance

SMS in rural areas. I think this is one of the biggest discrepancy i see between our field work and the belief in the community.

I also heard all over the place that sms is not a barrier. but in practice it is a major barrier. We couldn't find one farmer in Mali and Burkina who has ever recei or send an SMS. Illiteracy is not the only problem. I'm just returned 2 days ago again from Mal where i i've talked to representatives of farmer groups who are perfectly literate but are not able to use SMS. Illiteracy is one of the 3 factors with language and interface that explains that.

It seems that there is a growing set of publications that starts to reflect what we are observing in our projects in India and west africa that shows that SMS is not a channel of information for farmers (see again the sources in my post). This is true for agriculture information, but this is also true for e.g. financial service like m-pesa in Kenya. Sometimes relatives, or kids are able to read sms, but this is not a reliable option.

Now about voice-based services, this is where the major opportunity is for the future. Definetly, any human-based call center setup is not a sustainable scalable option. The only way to go is through automatic IVR system. There is a need for further research in that domain, and this is exactly what we are currently doing, but all the data show that if we can empower farmers directly to search and find information they need or feed the system with information they have (e.g. production, quantity, price etc) through voice technology, then this would solve current issues. IVR systems has been tested in the past but with a more traditional western approach of IVR system that are designed for people living in a technology-aware environment. Such approaches have miserably failed for people such as farmers. There is now a new movement and a new community developing new innovative approaches in that area which looks very promising. See for instance the report of the workshop we organized in Tanzania last june of voice-based services.

In terms of examples, we are experiencing in different porject in India and Africa such approach, but this is still in development. See e.g. VBAT (Voice Browsing Acceptance and Trust) or W4RA (Web Alliance for REgreening in Africa) or VOICES (VOIce-based Community-cEntric mobile Services for social development)

I would like to say that in any initiative, new challenges usually emerge and requires that they are adressed as they come.For the case of M-pesa in kenya,people who are illeterat e still use the service of sending and receiving money.When they receive money ,the senders usually call them to inform them that they have send a given amount to them. They then look for mobile operation agent whom they trust would not cheat them.In the case of farmers.Services given to farmers can be having accounts to accomodate the unbanked population as well as making transations easier as they can make order of inputs using the mobile service   

Hillary Miller-Wise
Hillary Miller-WiseTechnoServeTanzania

To pick up on Steph's comment, I agree that data-rich content on feature phones as well as voice-based services are more promising for achieving impact than SMS. The question I have is how economically sustainable voice-based services are in this market. IVR and call centers are clearly the most costly information delivery mechanisms. We also know that farmers' willingness to pay for agronomic information tends to be low. What indications do we have that willingness-to-pay will increase to the point of equilibrium with the cost of these services? Or do we think that the costs of voice-based services needs to be covered in some other way?

Hillary
 

Benjamin Kwasi Addom
Benjamin Kwasi AddomThe Technical Centre for Agricultural and Rural Cooperation (CTA)Netherlands

Yes, farmers' willingness to pay for agronomic information may be low because of the risk involve in investing in farm activities by rural farmers, and also the decades of free services from national extension services. But to be optimistic, the trend is likely to change in the next few years where a business-oriented farmer takes up insurance for his/her farm and therefore be more willing to pay for such services. Also farmer cooperatives that I talked about in my earlier post!

I think we also need to explore some models such as volunteers providing agricultural information or the use of national service or youth service in some parts of Africa to smoothly transition into a more sustainable economic models. Africa especially has rich resource of youth who are skilled in ICTs and also in agriculture. Preparing them through college education into the national service or youth service could be an asset for sustainability. This will be a longterm investment though i.e. through their educational programs to have some background in development or rural sociology in order to handle the task.

These approaches may be used to cover the costs associated with voice-based services that Hilary is anticipating, in the short term.

Ben

stephane  boyera
stephane boyeraSBC4DFrance

Hillary's post has a couple of very important points:

*IVR cost

*willingness to pay at the farmer level

On the IVR cost side, I think that people are mixing two parts that are unrelated. From the service provider perspective, it is very cheap to setup a (low-scale) voice-based service platform. Operating the platform is also very cheap in the sense that content and actions on the platform by the service provider is usually done over a traditionnal web browser link. IVR technolog is not costly at all, at the same level as other technology i would say. That said, what is very costly in most setup that are deployed today is the human-handling of the calls, either in a synchronous (call center, a human answer a human) or asynchronous (the caller drops  a message than is then manager by a human) mode. This is the major bottleneck today, that the technology could easily fix in a very near future.

On the willingness to pay, this is a key factor, related to the previous paragraph. An ivr platform is very cheap to operate if airtime is paid by the caller (the farmer) or is free.

Let me start with the 'free' option. It is not a joke, it is a trend that we are seeing all over the world where intra-network call are getting close to free or are free. In lots of developed countries, fixed and mobile airtime is now free. e.g. All operators are offering through their fixed-line combo free call to all mobile and fixed phones in france. This trend is expanding in developing countries too. In India, reliance at least is offering free call within its network. In tanzania, one operator (can't remember its name now) is now offering vouchers that allow unlimited call in the network for a given duration (1h up to 24h) at very low cost. I'm convinved that this trend will expand.

Now specifically on the willingness-to-pay. I've to say that this was also a big discovery for me, as i thought this would be a major problem. I've to say that i was very surprised in our operation in Sahel (Mali, Burkina) and India (UP) to find out that this was not a barrier at all. People are very pragmatic, and while it takes time to integrate a new service, then when people trust it, they have no problem making the relation between price and value. For instance, in rural areas, most of farmers have absolutely no way to get any support outside their circle. The only source of information is the radio. Therefore, the airtime is not the problem if this is to save (some of) their production. For me, the eye-opener was the use of airtime for entertainment: all community radios have broadcast that allow people to call in to vote for songs, or say a message, and those are very very popular, tons of people are calling in.

My little experience in few countries lets me think that the problem is not the airtime, the problem is not paying more than airtime, the problem is the return on investment. People pays for service that brings more to them than the cost (saving costs, saving travel time, increasing income, etc). Few very specific stories i had from last week was as simple as women producers beeing so happy to be able to stay at home and look after their children while being able to phone to see if there are customers for their products, rather than going to the market and visiting all the usual buyers. They were spending up to 10.000CFA airtime (15eur more or less 20$) in the month mostly for that.

steph

Benjamin Kwasi Addom
Benjamin Kwasi AddomThe Technical Centre for Agricultural and Rural Cooperation (CTA)Netherlands

@Stephane - what is very costly in most setup that are deployed today is the human-handling of the calls ... the major bottleneck today, that the technology could easily fix in a very near future.

If I really understand the above statement, then I will disagree with you, Stephane! Correct me if I mis-understood your statement.

I believe the human aspect is key and remains important in information provision for agricultural "production" - market information may be possible.

I worked as agricultural extension officer in a developing nation for over 5 years. With that experience, I will tend to differ from aproaches that aim at replacing intermediaries with ICTs/mobile technologies. There is the need to improve the brokering process rather than removing it completely. ICTs are "communication technologies" to support human activities. For example radio messages may be used to create awareness, signal, and disseminate information on new innovations/technologies but may not be appropriate for providing the technical know-how or training in the use of the innovation.

We see TV commmercials in the US that are returning to "human focus" services rather than automated messages.

If the cost in handling those call centers is an issue, we will need to see how we can manage that rather than trying to "fix" them purely with technologies.

Ben

 

stephane  boyera
stephane boyeraSBC4DFrance

Ben,

you are making a very valid point and that surely need further discussion. I totally agree with you about the fact that human are critical. I also totally agree that there are informatin that would need more than audio or sms message, but will require physical training on the know-how.

That said, i believe we all agree that there is a huge needs in terms of information service at the farmer level. I'm sure you will also agree that given the number of extension agents in the field, the more we can save their time, the better they can serve farmers. So all the points, imho, is to remove people where they are not critical, so that services can increase in quality, quantity, and efficiency.

So my reasoning was the following: we are having a discussion on how technology can help farmers to receive information that don't need to be conveyed by extension agents. SMS is a barrier for lots of farmers, because of the litteracy/interface/language challenges. So let's use another technology, voice, to deliver the same info, but with the same objective as having an automatic system.

Because if you have people in call centers, that can handle up to 70 requests a day, I'm not 100% sure that you have improved the extension agent model at all, and that people in the call center who not be more efficient in the field as extension agents.

Steph

 

 Mira
MiraUnited Kingdom

Hello Stephane,

Many thanks for your insightful contributions to this discussion!

I just wanted to comment on your assertion that a call centre with handling capacity of 70 requests per day may not be an improvement on existing in-person extension service delivery.

As a matter of act, I have been working on some research on the efficiency of extension service delivery and I find that extension agents in Ghana are occupied by the delivery of technical advice roughly 15-20% of the time. Administrative and project delivery activities are prominent among their activities.

Certainly, it would depend on the level we are considering (district, region, or national) but responding to 70 technical questions per day strikes me as a considerable capacity. In many cases I would expect it to provide efficiency gains over in-person delivery. Yet, what I would be concerned about is the articulation of the demand for the service. From what we are seeing, in-person technical advisory services are triggered more so by the extension agents noticing the need and delivering the advice, rather than farmers expressing the demand.

Cheers,

Mira