This is a guest post by Gonzalo Lezma from Amazon Mexico FP&A.
The Financial Planning and Analysis (FP&A) team in Mexico provides strategic support to Amazon’s CFO and executive team on planning, analysis, and reporting related to Amazon Mexico. We produce and manage key finance deliverables, such as internal profit and loss (P&L) reports for all business groups. We are also involved in planning processes, such as monthly forecast estimates, annual operating plans, and 3-year forecasts.
Our team needed to address five key challenges: manual recurrent reporting, managing data from different sources, moving away from large and slow spreadsheets, enabling ad hoc data insights extraction by business users, and variance analysis. To tackle these issues, we chose Amazon QuickSight for our business intelligence (BI) needs.
In this post, I discuss how QuickSight has enabled us to focus on financial and business analysis that helps drive business strategy.
Fully automating recurrent reporting
Creating and maintaining reports manually is time-consuming due to dense data granularity and multiple business groups and sub-products. This involves a lot of data to process and numerous stakeholders to please. Therefore, recurrent reporting requires allocating human hours to elaborate those reports, check the spreadsheet formulas, and rely on attention to detail to validate the numbers to report.
QuickSight dashboards show the Profit and Loss (P&L), which update as soon as databases refresh, simplifying recurrent reporting dramatically. There is no need for human intervention, which eliminates any risk of error during the elaboration of recurrent reporting. The maintenance and elaboration time has decreased from a week to zero for processes that are currently in QuickSight. We employ the QuickSight alerts feature (as shown in the following screenshot) to remain informed when specific metrics exceed a predefined threshold. This enables us to stay cognizant of significant changes in our P&L at a granular level.
Data from different sources
With new marketplaces and channels constantly emerging in Mexico, not all of them are integrated into the financial planning system; therefore, shadow P&Ls and reports are common and unavoidable. Therefore, the team has to find ways to track them without compromising accuracy or consistency, which poses significant additional challenges. Moreover, with multiple channels and teams reporting those numbers, it’s time-consuming to manually update the data source from every team we work with.
QuickSight can onboard data on channels and products that are relatively new and haven’t been onboarded to official planning systems. The team has numerous options to load data, including Amazon Redshift, CSV files, and Excel spreadsheets. There is virtually no limit on the granularity and scope of our reports.
Large and slow spreadsheets
Although spreadsheets are a popular tool for financial analysis, they have limitations for large and complex datasets. This affects performance, reliability, and validation. Spreadsheets become slow, bulky, and prone to errors, making it challenging to manage large datasets efficiently.
The SPICE (Super-fast, Parallel, In-memory Calculation Engine) in-memory engine that QuickSight uses is unparalleled, compared with other solutions the team tried in the past such as Tableau and Excel, eliminating the need for large spreadsheets dramatically. In addition to the time spent in elaborating the reports, the team was having a hard time reading and visualizing them.
The MX Financial Planning and Analysis Dashboard shown below shows the main contributors to the Gross Merchandise Sales for our business. If sales growth is at 20.92% as it says in the graph, we know that 9.09% is due to our NAFN channel. The graph at the bottom shows which products drove the sales increase.
Ad hoc data insights extraction
The finance space frequently requires ad hoc financial data on recent historic trends for a particular product and timespan. Given the number of channels, products, and scenarios the team works on, this creates a big problem to tackle. Extracting these data insights requires significant bandwidth, which can take away from other essential tasks the team needs to focus on.
Amazon QuickSight Q can answer any simple question about the data in a straightforward and nimble manner, allowing the team to handle ad hoc data insights using natural language requests. The following screenshot shows a graph we frequently get using Q to report shipping costs.
Providing accurate and insightful variance analysis is a significant challenge for anyone working in financial analysis (for example, explaining price or profit per unit by separating mix effect and rate effect). Huge and difficult-to-understand spreadsheets might sound familiar to anyone who has tried to tackle this problem in the finance space.
With QuickSight URL actions (as shown in the following gif and screenshot), the team can right-click on the variance they want to dissect and link to another sheet with granular detail that has a decomposition of the main drivers explaining that particular variance, replacing the huge and cumbersome Excel variance analysis tool the team used to have.
All in all, the dynamic and interactive nature of the dashboards allows our internal users to go deeper into the data with just a click of the mouse. Now, building visualizations is intuitive, insightful, and fast. In fact, the whole solution and tools were built without the need of a dedicated BI team. In addition to this, we developed internal QuickSight dashboards to view our own customers’ QuickSight usage, so we have perfect visibility on which areas and users are more active and which features are more used by our partners.
With our QuickSight solution, we have automation, self-service, speedy reaction to requests, and flexibility.
To learn more about how QuickSight can help your business with dashboards, reports, and more, visit Amazon QuickSight.
About the Author
Gonzalo Lezma is the Mexico Finance Manager for the Amazon LATAM Finance Team. He is a lifelong learner, tech and data lover.
This is a companion discussion topic for the original entry at https://aws.amazon.com/blogs/big-data/amazon-mexico-fpa-dives-deep-into-financial-data-with-amazon-quicksight/