Myer unlocks fresh business insights with Google Analytics 360 migration from USA to Australia

As one of Australia’s most iconic department stores, competing across many retail categories makes business intelligence critical to Myer’s success. With online revenue growth of 26% in 2019, myer.com.au is their largest store and most critical pathway for ongoing growth.

To drive this growth, accurate reporting and insights into how customers navigate the online store is crucial.

Unable to aggregate sales and loyalty program data

For the past three years, Myer has been using the Google Analytics 360 Suite (GA360) to gather business intelligence. The business is accumulating a wealth of historic interaction data – at a rate of around half a million interactions and 30GB of new data a day. 

Much of this data was stored in the US, with queries being run by a single compute instance. Because of this, the data was inaccessible to the onshore Google BigQuery-powered data hub that Myer was building to aggregate business intelligence from multiple sources. 

Myer was therefore unable to produce detailed insights by combining online interactions with its wealth of other sales and loyalty program data.

As Myer has been a long-time user of both Google BigQuery and GA360, around 80TB of GA360 data in the US had been accumulated.

Additionally, daily GA360 queries were written in LegacySQL and were being run by a single instance which users had limited experience with. This made the queries and the downstream dashboards difficult to maintain and update. As a result, users were not empowered to resolve issues themselves. Instead, they had to work around issues or call in third-party support.

More efficient queries

Two key needs had to be addressed to overcome these challenges:

1. Migrating the data from the US to Australia to combine it with the existing data hub
2. Executing the scheduled queries more efficiently.

To migrate large datasets from one region to another, two approaches have been published by both Google and former-Shine employees. Both involve exporting the data to Google Cloud storage and re-importing it to BigQuery in a dataset that is in the correct region.

Instead of these approaches, Shine made use of the Google Cloud Platform’s Data Transfer service.  This allowed for the automatic migration and synchronisation of datasets – while also providing Myer with a cost-effective means of moving 80TB of data.

For the daily queries against the data, the queries were upgraded to Standard SQL and executed as Scheduled Queries within BigQuery. This provided Myer with a readily available, clear view of which jobs they had running and when they were scheduled, all from the Google Cloud Console. In addition to productivity improvements, it also removed the need to have a compute instance running 24/7, with its ongoing patching and maintenance overheads.

Unlocking fresh insights and opportunities

By combining online store analytics with other sales and loyalty data, Myer has unlocked fresh customer behaviour insights and opportunities for sales growth. The business has also established a firm foundation for its team to develop additional queries and insights. 

Further, during the migration of the queries from LegacySQL to StandardSQL Shine helped Myer to identify several incorrect queries which were producing misleading results. This has increased Myer’s confidence in both reporting and understanding the data.