Through supporting MNOs to deliver projects under the mFarmer and mNutrition initiatives, the mAgri team have generated a number of insights and learnings on how MNOs can better reach rural populations. A target market which combines price sensitivity and often lower digital literacy, requires careful service design and compelling, localised content as a precursor to subscriber acquisition. Potential subscribers’ initial experience of being on-boarded to the service can be an important factor in how far along the customer journey (Figure 1) they travel and whether their interaction moves beyond simply trialling the service to becoming a revenue generating repeat subscriber.

 

Figure 1:

 

Our existing grantees’ go-to-market strategies can essentially be spilt into two types of model, field agent-led versus self-registration. Each model has their specific benefits and drawbacks, and are able to influence each step of the customer journey to varying degrees (see table below).

Agent led models:

By agent-led, we refer to ‘feet-on-the-ground’ field agents who are generally mobile and therefore able to move locations in order to promote the service to different communities and increase subscriber acquisitions. This differs from the typical airtime or mobile money agent operating from fixed territories or premises respectively.

Agent-led registration is particularly well suited to educating the customer about the service’s value proposition, registering and on-boarding of new users; important if the service is primarily focused on user acquisition. However with services designed for a rural population we have seen that all but the simplest of services do not scale very quickly using this approach. Agents need to spend time with potential customers not only to explain the value proposition, but to profile customers in order for targeted (or personalised) information such as crop-specific farming tips, market prices and local weather forecasts to be delivered. Such an approach is lengthy and for agents earning on a commission-only basis (for the new subscribers they acquire for the operator), the incentives tend to be linked with acquisition of high numbers of users.

For agent led models, the significant investment in recruitment, training and oversight in addition to the sheer number of agents required for the service to reach scale can make the overall cost of acquisition prohibitively expensive. Bharti Airtel’s Green SIM service in India, developed in partnership with IFFCO, was able to overcome some of the barriers to establishing a new rural agent network, by leveraging the co-operatives’ existing network of agri input retailers with whom the farmers had already developed trust relationships. In December 2014, Green SIM had over 3.1million active users and a network of 2,000 agents. Compare this with the instance of an unknown agent appearing in a rural community for the first time, and it highlights the agents’ uphill and time-consuming struggle to achieve new sales.

If agents are not permanently based in any one single area, then problems can appear with the latter stages of the customer journey, where subscribers are perhaps seeking help with the service, or clarification over how much the service costs to re-subscribe or indeed how to re-subscribe. We have seen issues with 30 day subscription services where the only interaction a user has with an agent is at the point of sale. In some instances, the farmer may not have thought about resubscription when enrolling to the service for the first time or the agent may have failed to fully explain the resubscription process. Even if the agent did explain it, it is common for users to simply forget the pricing policy or resubscription process.

With low cost services, the incentive structure should be carefully balanced across the collective activities of registration, profiling and repeat usage of subscribers, to avoid agents simply take the path of least resistance in order to achieve a sale. Without agents properly on-boarding new subscribers, the user’s experience is compromised and their likelihood for resubscription is greatly diminished.

Self-registration models

At first glance, the cost of acquisition per user under self-registration models appear to be much lower than an agent led model, however the upfront cost of promoting the service, with an unknown reach and conversion rate, can become prohibitively expensive if the effectiveness of marketing campaigns are not well managed or monitored.

While self-registration can overcome some of the operating costs associated with agent-led models, they can also run into problems with repeat usage in the absence of face to face on boarding and education of the service at the point of initial subscription. New subscribers are expected to use their own initiative to explore the service and this can be problematic from a number of perspectives, not least the digital literacy of the target population. Integrating ‘sign-posts’ for accessing help into the service are important so that users can access support if required. We have seen with our current grantees that users self-registering, predominantly use less than 20 per cent of the options available to them in the service. A lack of product knowledge or understanding, lack of clarity on pricing and worries about battery life have all been cited as reasons users don’t ‘explore’ the full service available.

Self-registration models benefit from having an unlimited rate of acquisition, as long as the technology behind the service can support a rapid growth of users without affecting customer experience. Additionally, services offering a simple value proposition and requiring little customer information for localised content to be delivered are most suited to self-registration. Airtel Malawi, with their M’Chikumbe 212 serviced delivered via IVR, only requires a user to register their primary crop during registration, with additional details (such as gender and location) requested after their third and fourth call into the service respectively.

Complementary strategies

To date we have seen greatest success with hybrid models; services which allow self-registration but with service ambassadors present in the communities of users. These ambassadors are not agents in the traditional sense, but stakeholders within the farming communities who identify the service as a positive tool for farmers to support them in improving their day to day farming activities.

It is worth mentioning that neither of the models outlined above should be deployed in isolation. A complementary above and below the line (ATL/BTL) marketing campaign to support both acquisition and retention is important to ensure the service remains at the forefront of users’ minds. Messaging through SMS, Voice SMS (out-bound dials (OBDs)) and service notifications can be hyper-targeted to different user segments to promote certain types of behaviour and increase rates of retention and resubscription.

You can learn more about our grantees’ registration strategies from our existing mNutrition initiative services across Africa and Asia by registering for our upcoming webinar “Solutions and iterations in mAgri products: Processes and interim findings”

Register your interest here.

Examples of how the models differ at various stages of the customer journey:

 

	Agent-led	Self-registration
Sales Channels / Marketing: Activities used to raise awareness of the service, drive interest and demand.	Able to explain full value proposition to face to face with users.   Rate of acquisition directly related to the number of agents available.	Reliant on ATL/BTL marketing activities to effectively explain the value proposition.   No limits to rate of acquisition.
On boarding: Activities conducted to get the user ‘set up’ (any KYC, registration or profiling related activities) and access the entire service offering.	Can collect registration information face to face and explain what the information will be used for.	Reliant on user trial, word of mouth.   Requires clear welcome messaging to on board new users.
Educating: Activities conducted to orient and educate the user on the service (also referred to as product literacy – knowing how to subscribe/unsubscribe, navigate content, understand tariffs, notifications, etc.).	Able to educate users on how to navigate the service and access different features.   Can be ignored by agents if incentive is focused purely on sales.	Users required to explore options for themselves.
Content: Describes all elements of the content (formats, channels, languages, frequency of updates, etc.) currently live and either pushed to users or available to be pulled (including agricultural help lines).	Can provide the user with examples of content at the point of sale to build credibility of the service.	Trust has to be built with the user to prove that content will be worth their investment (both time and financial).
Payment: Activities conducted to collect payment (cash to agents, airtime deductions, etc.).	Able to explain payments to new subscribers, however if this involves a 30 days cycle then it is possible that farmers will forget what and when dues are necessary and how to resubscribe. Can be mitigated to some degree by using notifications channels to update the farmer as their subscription bundle approaches the renewal date.	Needs clear messaging when users register, or as users approach a payment moment (e.g. usage threshold or renewal date).
Getting help: Activities conducted to get help on how to use the service (customer service).	Can demonstrate how to access support at point of sale (e.g. dial helpline) however, user may not remember how to access support at the point at which it is required.	Users need to be ‘sign-posted’ to help options.
Notifications and retention: Describes all elements (formats, channels, languages, frequencies) of all notifications (subscription confirmations, expiration notices, etc.) and activities or experiences for keeping customers on the service.	Extremely important complimentary channel to both models for informing users about new content, promotions, renewal dates and service status.