Case Study: 
Seeling Construction Materials in the Amazon

This research was developed for a retail company that has been operating for 80 years in the Brazilian Amazon territory. The company's great strength is a triad of efficiency in logistics, connectivity and installment plans. The company started new sales operations with construction materials and wanted to understand how a better experience of buying these products in a digital context could be, as well as what would be the best ways to expand the business based on these customer profiles.

Context

Progress

Results

My role as a researcher was to align stakeholders’ expectations and available resources, as well as allow them to clearly visualize the scope of the research with the limitations of the context (budget, sample, time). I worked in partnership with a teammate, reconciling this activity with other ongoing research.

The survey lasted 4 weeks, comprising a week to study and understand the best recruitment criteria, a week to carry out data collection, a week to finish the collection and start the analysis and the last week for the handoff reports.

The research was demanded by the company's e-commerce managers, however, the people of greatest interest were those who were dealing with the curation of the products - that is, those who were selecting the items to be sold and managing the stock and delivery. Consequently, the research needed to deliver inputs that clarified what to offer in terms of products, and how to operate logistically in the best way.

In the recruitment phase, we chose to study the best selection of the base from the analysis of its sociodemographic and buying behavior aspects, in a quantitative aspect, with the help of a data analyst. This helped us get an initial picture of buying behavior and profiling across a large sample. We selected a base of customers who had purchased in the last three months to approach them initially with a questionnaire.

The questionnaire initiated our qualitative approach, aiming to obtain more detailed information about some profile characteristics as well as satisfaction feedback regarding the purchase made. Through it, we also established a way of communicating with those who were interested in participating in the interview phase. The call to customers was made by sending SMS to the registered numbers.

Interviewed people were those whose sociodemographic, consumer behavior and questionnaire data we already had. Our questions explored more attitudinal aspects, relating the data collected to their pain, needs and motivations. The interviews lasted from 15 to 25 minutes. We also conducted a focus group in a municipality near Manaus, to gather more information about one specific profile.

Respondents' adherence to the survey was catalyzed by the offer of rewards. A credit bonus was offered that could be spent at the company's stores. This reward provided an index of respondents to the questionnaire of more than 12% of the base.

The analysis process consisted of building a narrative of findings that was structured in a layer of knowledge about the client, acquired from the process of studying the best way to recruit. We separated behavioral, attitudinal, and sociodemographic aspects, and gradually the profiles were distinguished in geographic contexts. Customers from the city centers (like capital Manaus) had different pains and needs from customers residing in the most remote areas of the Amazon. Two profiles were emerging.

We chose to build a profile narrative with persona features, but not necessarily adopting their exact narrative form. Our intuition was that research stakeholders needed information described in a more instructive and less narrative way.

The survey results made it possible to calibrate the curation of products to the demands of the clientele, increasing sales and optimizing stock. The logistical expectations of each profile were different, and this was also adjusted, resulting in the optimization of the operation's costs. Still, the research results provided several other insights of other products that can be marketed to the most remote areas of the Amazon.

The next steps of the research foreseen included refining the understanding of what the customers from other parts of the Amazon want, as well as focusing on how to customize the shopping experience of the company's website to the specific demand of construction materials.

The main difficulties resided in establishing an internal flow – still inexistent – ​​to operationalize the sending of calls for research on a broad basis, through SMS. Also, it was the first time that the company was rewarding survey respondents and, therefore, this flow also needed to be pioneered by us. Such processes ran more slowly, but successfully. For future research, we would probably need to add more time to build new research-related workflows not yet implemented in the company.