Improving Services in Low Income Markets
For businesses selling products and services at the base of the pyramid, understanding low-income consumers and the choices they make is critical to meeting customer needs, boosting sales, and delivering real value.
Inclusive business Dr. Consulta run a network of medical centres that provide primary and secondary care in São Paulo, Brazil. With over 85,000 visits a month they provide affordable care to 250,000 people at a fourth or fifth of the price of alternative treatments.
I spoke to Jorge Tung, Chief Product Officer, who told me how they collect invaluable feedback from their customers and are able to use it to better address the needs of their patients.
Who are your target customers and why is understanding them important to your business?
In Brazil, the healthcare service is free, but it’s quality is bad, with patients waiting 3-4 months for an examination. We sit in between those who cannot pay for health insurance and those who have it, about 100 million people in Brazil. It is a big enough market, the challenge is serve it affordably, accessibly, and efficiently, so we rely on our understanding of the customer and technology, to tailor the service. We have to make the process extremely lean.
First, we try to understand our customer, build some insights, or try to understand a problem and try to figure out a way to solve it and deliver best customer service.
So, for example, if we need to understand what are the motivations for people going to our medical centres, we’ve found three types of patient. Through customer insights, we see how we should treat these three types of people, what type of communications should we send them, how we should treat and engage each one and so on.
Talk us through how you collect this customer feedback.
We have many tools for getting feedback, mostly from patients. The first are SMSs to all patients after every visit. We use Survey Monkey on a monthly basis for getting more detailed information.
We worked with Acumen Lean Data as a part of our customer discovery using a 2-fold research approach. First was qualitative research- we sent a psychologist in the field, to experience the whole process. She pretended she was a regular customer, she went through the process and looked at the issues that came out. She then interviewed our customers in the medical centre in a series of questions in a 20-30-minute interview.
Then, we built profiles of customers, assessed customer journey and created some hypothesis. After that, we engaged with Acumen Lean Data to get quantitative information to validate some ideas and answer- are the hypothesis correct?
For example, we believed that the physician was very important to the relationship with the patient. This proved not to be exactly true, it was only true for a part of the population, the ones that had a chronic or ongoing disease. It is important for them to build a relationship. Other regular customers prioritise availability and relationship with the physician equally.
How often do you do this?
So, for our three types of research:
As we are growing fast we understand that our customer base is morphing over time. Last year we had less than half of the medical centres we have today. As we are expanding, the profile of the new patient changes, so we should continue doing research and surveys.
Are there any challenges to collecting that data and how have you overcome them?
There are a lot of challenges. For longer surveys, we offer prizes. We sent around 50,000 invitations and only 1,500 came back. For text messages, we have an engagement of around 20%.
Cost is also very important; we use the available technology to do it as broadly and quickly as possible. Only on in depth research can we afford to interview customers.
How much of a role does technology play in how you collect feedback?
Throughout the entire process; recruitment, engagement, how we send surveys, collect data, how we analyse. We are still developing the strategy, so there are patients, for example, with chronic disease that complain they receive too many text messages. The other issue is the data that we analyse, we understand that many people mistake the keys answering SMS messages. Many people rate their physician 1, when they want to send 10, there are also lots of human errors.
How do you use the data to ensure it is feeding back into the business and worth doing?
We are still trying to work it out. For example, customer satisfaction for physicians are part of their reward scheme. We compile feedback and send it to the physician, so they can evaluate and see how they can improve, and when they improve they are rewarded for it.
The yearly research, we feed back into things like building the website and mobile application. We are trying to build a website that can serve most of our patients, but we understand that some don’t have access to technology, some don’t book services online. For example: it’s very common in Brazil that the wife schedules an appointment for her husband, so 70% of visitors on our website are female, so we have to build the site more towards their needs.
Do you have examples of how you have improved the service for customers?
Through the research we understand which features we should improve first. For example, helping customers understand we are real! Many don’t trust us, they say: your service is 20% of a real one, it’s too cheap, you are probably fake or low quality. How can we build trust with patients? We publish physician profiles on the website to bring a more personal touch and bring us closer to the patient.
What advice would you give to other inclusive businesses who are looking to setup feedback systems to understand their customers?
Before you try to setup any surveys or feedback systems, sit with the customers and talk to them in person. Experience the service you are offering, get some insights, and build your hypothesis. If we hadn’t done that before engaging with Acumen Lean Data, we would have done a very long questionnaire, looking to answer all our questions and once we got the results, we wouldn’t have been able to make sense of the results. So, build the hypothesis first, then look to prove or break them with data afterwards.