The Retail Trade survey has been collecting online sales since the March quarter 2013. The data was previously published as an experimental series, in original data only, as an Appendix to the Retail Trade publication. It was disaggregated by whether the retailer was "Pure-play" (online only) or "Multi-channel" (mix of online and physical stores).
The Pure-play and Multi-channel split will remain available on request but will no longer be published in the Appendix on a regular basis.
Over time, the split between Pure-play and Multi-channel remained stable, with Pure-play online retailers averaging 38.3% of total online sales.
In September 2020, Pure-play retailers made up 31.3% of online sales and 3.3% of total sales. Online sales for multi-channel retailers made up 7.3% of total retail sales.
From July 2020, the Online series will be published with a grouped industry split; a Food group (including the Food retailing, and Cafes, restaurants and takeaway food services), and a Non-food group (all other industries). Due to the limitations of online data collection, a finer split by industry is not possible at this stage.
The online series continues to use the same source data, and represents purchases made via the internet from employing retail businesses who predominately sell to households. The series excludes direct imports (e.g. purchased directly from an overseas website) and sales from 'households-to-households' through third party websites for example. More information can be found in the information paper Measurement of Online Retail Trade in Macroeconomics (cat. no. 8501.0.55.007).
Online sales, including click-and-collect, have become increasingly popular during the coronavirus pandemic. It is worth noting that retailers have had to adapt sales channels very quickly during the pandemic, and it is possible that the value of online purchases has been under-reported since April 2020. Importantly, total retail sales will not be under-reported.
Seasonal adjustment has been calculated for the new series using the concurrent seasonal adjustment method, meaning that seasonal factors are re-estimated each time a new data point becomes available. Unusual real-world events, such as COVID-19, can distort estimates calculated using the method. The Online series remains experimental, and caution should be used in interpreting the results. Like the total retail series, trend cannot be calculated due to the volatility of the retail series during COVID-19.