Angie—Fu

Airtel Cut the Line

Project Background

IDEO.org partnered with Airtel Tanzania, one of Tanzania’s largest mobile network operators (MNO), to use qualitative and quantitative data to better understand, and subsequently design more informed, human-centered tools and services for their expanse network of mobile money agents. Our team consisted of a project lead and design researcher, a business designer, a communications designer, and myself as an interaction designer, in partnership with Driven Data and the Bill & Melinda Gates Foundation.

Services

Remote research
Interaction design
UX design
Interactive prototyping

Theory of Change

Based on data and analysis of Airtel’s agent network and customer transactions, we’ve seen that the more a customer interacts with the same mobile money agent, the individual transaction volume increases at a strong, steady pace. We took this to indicate that frequent encounters lead to trust in the agent—and that trust potentially enables a larger trust in the mobile money system. Larger trust in, and exposure to mobile money enables MNOs to provide more accesible financial services to low-income customers who formal financial institutions aren’t paying attention to.

Design challenge

While we had some directional evidence, we needed qualitative feedback to better understand how we could further strengthen this relationship with a design-led approach. We asked ourselves:

How might we equip agents with the right tools and incentives to accelerate positive customers encounters and improve trust in mobile money?

Brainstorming incentives

Approach

We started by establishing some key questions that would help us quickly identify the critical, unmet needs of both agents and customers. We learned that the most frequent issue with mobile money was that customers were too often running into technical issues like being locked out of their accounts—and too often agents didn’t have the right training, or enough time to solve the problem.

What if we could both train and incentivize them to provide a meaningful technical support touchpoint to customers? Based on data, we had a good idea when customer traffic typically slowed down and agents were unoccupied, so I worked an SMS-based service that would point customers to nearby agents had low customer traffic.

Challenges

There are significant considerations that make designing for the developing world very different compared to, say, San Francisco. Smartphone penetration is low, so feature phones and SMS's are the lowest common denominator. However, messages need to get straight to the point as only the first 140 characters of any message get sent. The rest of that long message sadly goes into the abyss, and dividing the text up unto multiple messages costs the customer twice the storage, and the MNO twice the cost.

Simulating with games

I worked with our communications designer to design several games that could help us understand the right incentives for agents, and frequency of communications for both actors.

These helped us shape the constraints and necessary moments of support that were built into our final prototype

Cut the Line

The Cut the Line program aimed at expanding agent into the first line of problem solvers for customer issues and concerns.

Customers are encouraged to approach agents to resolve common issues, such as PIN resets, SIM swaps, transaction reversals, etc. Agents will be equipped with trainings and tools to successfully address these issues, and also be rewarded for successful issue resolution.

When the support moment is marked as done, customers will be requested to rate their experience to ensure that the issue was successfully resolved, triggering an incentive for the agent.

Working hypotheses

Lowering the bar to reset PIN will lead to less customer turnover
Agents have potential to add value to customers beyond cash in and cash out
When an agent helps a customer solve a problem, the customer will return to that agent for more business

Agent training guide

Partner handoff

I led the charge of packaging the program details, technical specifications, and interaction diagrams to begin the implementation phase with our developer. While we were able to set up the basic tech architecture and start segmenting agents and customers in preparation for live prototyping, our funding fell short and the project has since been stagnant. It's a bummer, but our learnings about the dynamic between agents and customers has supported our projects in other East African countries.