With the rapid growth and traction of ShopBack, ShopBack’s data team ran into a problem—the number of data requests were going up as the company grew, and was taking time away from the team’s core data work.
Jacob Eyo, Data Analytics Lead at ShopBack, used to have to manually write queries and extract datasets from their databases, before emailing the spreadsheets results to his business teams. Daily and weekly reports were prepared manually, and there were days taken up entirely by attending to over 8 data requests. While a simple request may take him 30 minutes to complete, complex requests could take him up to 4 hours to prepare and complete.
Additionally, storing and managing query scripts began to be cumbersome. Most SQL queries used to extract the report result-sets tended to be similar with minor parameter changes, however, storing these scripts into saved repositories did not provide the context on what the scripts were developed for, and other necessary metadata about the purpose of the queries.
The data team tried storing these script definitions and descriptions in a corporate wiki, which included the script description together with the SQL syntax in a single place. However when ShopBack expanded into different countries, the team found themselves navigating and scrolling through long continuous pages of similar SQL queries, with overly specific syntaxes for each country.
The ShopBack data team needed a better way to manage, process and deliver the reports to their business users. Their business teams wanted more up-to-date access to the data needed to make timely decisions.