Deconsa (M)
Client goal
The Gates Foundation is a private organization with vast ambitions; one of its goals is to reduce extreme poverty worldwide. In this case study, the foundation asked McKinsey to design a basic financial-services offering for inhabitants of remote communities in Mexico.
The key word of the "client goal" is "design". So, this case is about designing a financial-services offering. Factors relating to desining a whole system are important to consider.
Description of situation
The majority of Mexico’s rural inhabitants are relatively poor, relying in part on government benefits for their livelihood. As they tend not to have bank accounts, they usually collect government benefits in cash from a limited number of state-bank branches. These branches are often a long way from where the recipients live, so it can take a lot of time and effort for them to collect their benefits. In addition, traveling to the branches can sometimes expose benefit recipients to crime along the way.
The Mexican government also owns and operates a chain of 22,000 stores, throughout Mexico, called Diconsa, which provide basic food, clothes, and other essential goods to the rural population. These stores are supplied via a network of central and regional warehouses and several thousand delivery trucks.
McKinsey study
McKinsey has been asked to investigate and assess the possibility of using the Diconsa network to provide a basic set of financial services to supplement the limited number of state-bank branches. The offer would start with dispensing government-benefit payments and move progressively to include savings accounts, bill payment, insurance, credit, and other financial products.
Question 1:
What should the team investigate to determine whether the Diconsa network could and should be leveraged to provide a range of basic financial services to Mexico’s rural population?
The description is long but it's very easy to establish a visual of what's going on, and that's why I recommend visualization as an important problem solving tool.
Question 1 is asked in an unusual way. McKinsey's first questions are usually about identifying key factors. However, this question can be re-written as "What key factors are important to consider in order to determine whether the Diconsa network could and should be leveraged to provide a range of basic financial services to Mexico’s rural population?"
From my point of view, the answers to Question 1 and the re-written question can be the same.
So, just imagine, to convert a network of stores to locations with financial service capabilities, how? What are the key factors?
The way I would think about solving this question would probably be: first, implementability. second, cost. third, benefit. Lastly, risks.
I think it's very natural to think about cost versus benefit in determining whether a project should be carried. But in the real world, one can't dive directly into cost versus benefit. That's why I feel I'd like to consider implementability first and consider risk in the end.
Under implementability, I probably would want to consider two main aspects. (1) coverage: how does the Diconsa network cover the benefit recipients? Ideally, we want a reasonable existing coverage in the sense that most recipients have at least one Diconsa store nearby. (2) Organizational capacity: managing a network of stores to distribute financial benefits may require a different organizational structure within the government. This requires the Mexican government have some degree of flexibility and will to change.
Under cost, I want to consider three types of cost: (1) Infrastructure: cost to equip the system with cash transportation tools like guarded cash trucks, and cost to upgrade Diconsa stores like adding safety boxes and hiring extra personnel. There may also be needs to build new Diconsa stores for the villages where there was no store before. (2) Administration: like previously mentioned, there may need to be a rather dramatic organizational change, implying cost related with labor, computer system, office space and etc. (3) Marketing and education: residents may be reluctant to utilize the store or use other financial services, so financial resource is probably needed to market the service and products to the residents, and to establish certain education programs for them.
Under benefit, I want to consider two major types of benefits: (1) Benefits to the recipients including but not restricted to reduced crimes, saved traveling cost, opportunity cost of labor as well as benefits that come with higher liquidity like more frequency dissemination of cash, and benefits that comes with the access to more types of financial products and tools. (2) Benefits to the state government including but not restricted to upgraded dissemination system, and positive spillover of this project to other government projects.
For risk, I think there are three main risks to consider. The first is security, for example, risk of cash trucks being hacked and security boxes at the store being stolen. The second is administration, for example, the risk of corruption and embezzlement, especially at the store level. The third is acceptance and penetration speed. The recipients of benefit probably have very limited experience with financial products and service, so they may highly resist this systematic change. In this case, the penetration of the project may be extra slow.
It’s been a long time I have first seen this case. I didn’t read the sample answer before I structured the case. If we read the sample answer, it has a different structure from what I had. It really doesn’t matter how one structures a case, as long as the key factors are identified and well organized, it should be fine.
Question 2:
The team has estimated that it currently costs a family 50 pesos per month in transportation and food to make the journey to collect their benefit payments. The team also estimates that if benefits were available for collection at local Diconsa stores, the cost would be reduced by 30 percent.
Twenty percent of Mexico’s population is rural, and of this number half currently receive state benefits.
If all families could receive state benefits at their local Diconsa stores, how much in total per year would be saved across all Mexican rural families receiving state benefits?
Your interviewer can provide you with the following information if requested:
You can assume that Mexico has a population of 100 million.
You can also assume that families in Mexico have four members, on average, and that this does not differ by region.
Question 2 is a quantitative question, and it is relatively easy compared with typical McKinsey quant questions. The general guideline toanswering a quant question includes (1) make sure you understand the question and have written down the key numbers correctly; (2) ask for missing information if they are necessary to make the computation; (3) let the interviewer know your plan first before diving into your calculation; (4) make your calculation communication-friendly and if possible, walk your interviewer through each and every step of your calculation; (5) don’t stop at the answer, but offer comments and insights.
One special note is that for quant questions, and any other questions in case interviews, there is never a one-recipe-fits-all solution. The general guidelines are created to showcase you a communication-efficient and friendly way.
I will not showcase the solution of Question 2 here because it is very straightforward.
Question 3:
The team conducted a survey on a sample of the rural population in three different regions of Mexico. These populations were given a number of statements about the concept of collecting their benefits at the nearest Diconsa store and asked how much they agreed with each statement. The average response to some of the questions by region is shown below:
What are your observations about this information, and how would you explain these trends?
Question 3 is a graph interpretation question. Like any other questions, the key is not only to structure your answer, but to structure your answer with priority. There could be many insights from this graph, but which insights to offer first and second can make a difference. Following the insights, it often is desired to comments on the potential drivers of the insights, and possible actions that your client could take.
However, before you do all the above, make sure you understand the graph. For any concepts you have doubts with, ask the interviewer for clarification. Remember that you can always ask questions and they seldom hurt you. For example, in this graph, I probably would ask what does “I am happy to collect my benefits” mean because does it mean that a person is happy to collect his or her benefit the old way or does it mean that a person is happy to collect his or her benefit the new way?
After the question is clarified, I probably would begin by saying that there are a few observations: (1) response vary quite significantly by region, with only one consensus formed over the reduced time to collect benefit; (2) Region A seems to be the most welcoming region with scores higher than the other two regions, while Region B is the least welcoming region. (3) Zooming into Region B where most of the response seem unpromising, it seems that people have security and trust issues.
After offering the observations, or as I offer each observation, I probably would link the observation to potential drivers and offer suggestions related to the eventual goal of the case. I probably would say that we may want to do a few pilots in the regions that seem welcoming, and in the meantime, look into the drivers behind the unwelcoming response of Region B. The potential reasons may include higher corruption and embezzlements at the Diconsa store level. For example, store owners may hold the benefits hostage for extra personal benefits.
Question 4:
In a meeting with representatives of the state bank, they express concern about how challenging it will be to offer basic financial products to the rural population.
One of the representatives mentions that these are regions where people are poorly educated and unfamiliar with banking products, and where there is a high degree of crime and fraud. He asks for the team’s thoughts about how the bank could sell these services effectively to this population given the challenges.
Question 4 is a marketing & sales question. Although the question is placed in a non-profit setting, it doesn’t change its nature. You should know by know what I would suggest…structure, structure, structure…I think the sample answer given by McKinsey is a bit misleading in the sense that it sounds like a laundry list. I think the McKinsey answer is trying to send a message that for brain-storming questions, the innovativeness of your ideas is important. However, from my point of view, structuring is the first step and then comes innovativeness. (not saying that innovativeness is not important).
For this question, I probably would structure the answer into three buckets: products, marketing, and execution. For products, I probably would include design, pricing and payment; for marketing, I would include pricing & promotions, education & demonstration, and partnership; for execution, I probably would include piloting, differentiated strategies, and feedbacks.
There is an extensive list of ideas in the sample answer so I won’t expand my structure here.