Everymundo dashboards provide data insights by analyzing real-time User Search Data collected from the customer’s internet booking engines (IBE) by Everymundo’s proprietary script. With business insights, customers can identify their most strategic routes in real-time, understand the gap between search volume and load factor to identify opportunities for fare adjustments, and monitor user engagement with their airTRFX Pages on airModules, URL, or Route-level. Rolling out large amounts of dashboards with high data accuracy became a top priority in Data Analytics because these standard reports are available to all customers.
Data Cleaning
On average, over 9 million rows of raw user searches and over 200 thousand rows of raw booking data are collected from customers’ IBE daily. To ensure the correct, non-corrupted and consistent data is stored, we will perform data Quality Assuranceaccording to emDataStandards when customers are migrating to production on the final domain, and we will submit tickets to developers for any issues detected.
Data Blending
Complex but insightful dashboards typically are built with blended data, which combines multiple data sources for additional information. To create a visualization serving all customers the best method of merging multiple datasets designed beforehand is even more important. How the relationship among tables is defined affects the details and key insights provided in one individual dashboard, the loading and calculating speed of the whole workbook, and the total workloads that we will bring to the server.
Template Workbooks
To roll out the standardized tableau dashboards with high efficiency, Everymundo template workbooks were created and published in a shared internal tableau folder. The template workbooks contain demo dashboards with a pre-defined form and structure, including the fully-constructed charts and tables, the insightful key metrics and calculations, and clean dashboards design and layouts following EveryMundo Brand Guidelines. Template workbooks help save the dashboards producing time and maintenance costs because they provide consistency and uniformity.
Replication Process
With the template workbooks in hand, the replication process becomes simple. Please see the general steps below:
- Download the template from the EM Template folder through Tableau Online
- Open the template workbook with the Tableau Desktop → Click on the “Data” tab → Connect to the data source → Replace the data source for a specific customer
- Extracted data source with customized SQL query: Follow the instructions to re-edit the query for the customer
- Extracted data that needs to be published to Tableau Server (several dashboards shared the same live data source): Check whether the necessary data source has already been published on Tableau Server
- If yes, connect to the data source directly
- If not yet, create the local copy → Re-edit the query for the customer→ Publish to Tableau Server→ Connect to the live data source on Tableau Server and replace it with the previous one.
- Connect to other data sources such as Google Analytics, Google Ads, or Google sheet
- Go through all of the views inside the workbook, debug the issue caused by data source replacement, and replace the customized signal like currency code and date format
- Review all the views and filters and make sure the dashboard is functioning properly
- Publish the data source (if needed) and workbook
- Schedule the extract time to refresh data and update the dashboard automatically
QA
Data QA is the most important and final step of the dashboard replication process. During the QA process, our QA Engineer helps the team perform the below verifications:
- Verify the KPI metrics shown on the dashboards to ensure the numbers make sense
- Verify the dashboards are rendering correctly in both tableau online and airTRFX Control, which is the backend platform that supports customers’ self-services in TRFX and displays data insights
- Verify the dashboards are displayed as expected on the customer’s portal
Depending on the dashboards, QA Engineers would also offer suggestions in dashboard design or dashboard annotation, which help us improve the user experience of our team products. If any issues are detected or any modification is needed, Jira tickets for dashboard creation are rejected by a QA Engineer. Data analysts perform further investigations to identify the root cause of the problems or perform any dashboard improvements.
The Data Analytics team at Everymundo works actively creating new dashboards for various relevant insights and maintaining the current dashboards for up-to-date features. Those interested in Everymundo dashboards can explore the demos published on EM Tableau Public profile. Data is modified to maintain anonymity. To access dashboards, please email [email protected].