Apply for a job at Olist's Marketing Analytics Team: http://bit.ly/work-at-olist-marketing
Olist is the largest department store of Brazilian marketplaces. We connect small businesses to channels without hassle and with a single contract. We have connected more than 4.1 thousand sellers to stores like Americanas.com, Ponto Frio, Walmart, Submarino, Mercado Livre, Amazon, among others.
Responsible for taking advantage of growth opportunitties, this team work close to data to create new lead capture tools, prediction models and advanced analysis.
This repository contains instructions to solve problems that are used to evaluate the candidate skills. It's important to notice that satisfactorily solving this test it is just a part of what will be evaluated. We also consider other disciplines like the capability of extracting insights from data, business understanding, data visualization skills, documentation and design.
- Follow the instructions of README.md (this file);
- Download the test data available on Kaggle. You may want to solve the problem online using Kaggle Kernels.
- Take your time to read carefully the data documentation and understand how it works. If you have any questions about the data, open a public discussion on the dataset page.
- Use the tool you like the most or are more familiar with to analyse the data and solve the problem:
- Kaggle Kernels for Jupyter Notebooks (Python/R);
- Colaboratory for Jupyter Notebooks (Python/R);
- Google Sheets;
- Microsoft Excel;
- BI tools (Power BI, Tableau, Qlik, etc);
- Any other tool you like;
- Apply for the position at our career page.
- Send your analysis by email (you'll get the address at the selection process).
- You might be asked to explain some steps of your analysis at the interview.
Completing the test and applying for the postion does not guarantee that you will be called for interview.
Olist has released a Marketing Funnel Dataset from sellers that filled-in requests of contact to sell their products on Olist Store. The dataset has information of 8k Marketing Qualified Leads (MQLs) that requested contact between Jun. 1st 2017 and Jun 1st 2018. They were randomly sampled from the total of MQLs.
This is real data, it has been anonymized and sampled from the original dataset.
You may want to solve different types of problems with this dataset. Choose one that best suits you and have fun! Bellow are some examples of things you might want to do, but feel free to use your creativity.
Notice that you are not required to do everything listed bellow. Those are just starting points, ideas to guide your work. Remember that quality is more important than quantity.
Language Requirements: We like to experiment new things! Feel encouraged to solve the test with a tool you like the most. We accept solutions on spreadsheets or presentations in pdf. But you are really going to touch our hearts if choose tools like Python/R to solve this problem. Showing your code/solution is a plus.
- Analyse the performance/results of our aquisition channels through the marketing funnel
- Link the data with the orders dataset and estimate the customer value for each channel
- Create KPIs and a dashboard that shows the current funnel status and provide some insights
- Find some opportunities to improve the area by using data. Which channels should we invest more? Wich ones should we stop?
- Find correlations between data and try to predict an outcome (closing a sale, for instance)
- Develop a plan to attract and capture more qualified leads (tool based marketing)
- Propose some tests to analyse and improve our performance
Or just tell us something we don't know. Which business value would you create for Olist with this data?