Online retailers yet to harness big social data

By | | comments |
(Image via

When your online purchase doesn’t show up, it’s likely you’ll turn to Twitter. University of Sydney researchers  and  analysed a global database of 12 million tweets to find the good, bad and ugly of customer service and it's clear few are keeping up.

AROUND THE world, social media is giving consumers far greater visibility into supply chain processes and problems. For better or worse, manufacturing locations and labour conditions, warehouses, warehouse systems, packing processes, logistics partners and their delivery drivers have increasingly become associated with a retailer’s brand image.

At the same time, retailers and their logistics partners have set up virtual helpdesks on Twitter. To work out whether these have been effective, we analysed data collected from Twitter over the past two years. And the answer at this stage is: only partially.

Our database currently contains more than 12 million tweets from companies based in predominantly English speaking countries and their customers around the world. The data is spatially distributed as shown in the figure below.

Spatial distribution of customer tweets.

Social data is being co-created by retailers, logistics service providers and end consumers as part of everyday business practice in the supply chain, with customers approaching retailers on Twitter at all stages of the e-retail logistics process.

This includes problems with online shopping carts, order processing delays, issues with tracking information, delivery delays, service provided by the relevant logistics/parcel shipping company, damaged packages/items, missing items and reverse logistics processes.

Parcel shipping companies are generally approached regarding tracking information availability, accuracy and clarity as well as delivery issues including delays, driver behaviour, missing/damaged packages, missed deliveries and pickup locations.

We found conversations tended to be of varying lengths with two tweets (one from the customer and another from the company) being the shortest length of a conversation. Longer conversations could involve other companies as well as other customers, and included everything from the good, the bad and the ugly.

The good

Good conversations included positive feedback for the retailer, the logistics service provider or both. These are generally short conversations. Good conversations are also those where the problems raised by customers are resolved by the companies over the Twitter platform and the positive outcomes are visible to followers of both the companies and the customers.

These conversations provide value to the customer and reflect positively on the brand. In a small number of cases both retailers and their shipping partners were found to get involved in resolving issues for customers. We think there is considerable scope for joint customer service to become common practice amongst retailers and logistics service providers.

The bad

Bad conversations are those where the problems raised by customers are referred to Twitter’s direct messaging service platform or traditional channels (email, phone numbers) and the outcomes are not visible to other customers, thus generating little value for the brand.

Our research suggests customers rarely return to Twitter to thank a company for a problem resolved via another channel. However, customers do use Twitter to complain about not receiving timely responses via either the social media platform or traditional channels.

Another issue was the infrequency of involvement of parcel shipping carriers when they were specifically mentioned in customers’ complaints to retailers. We think the problem with this lack of involvement from shipping companies is the message it conveys to the consumer about the strength of the supply chain relationship.

The ugly

Ugly conversations were those unintentionally initiated by promotional tweets that reminded customers of recent or current frustrations with a retailer or logistics provider’s services.

These trigger the sharing of experiences between multiple customers and often place customer service staff in damage control mode. Consumer experiences with delivery delays around Christmas and confusion over information available on relevant websites in relation to order processing and shipping times contributed to the backlash.

Customer service conversations = big social data

A large volume of social media data gets created on a daily basis from these customer service interactions. Companies need to be examining both the volumes of unstructured social media data created by their own processes as well as by their competitors for a better understanding of necessary process improvements.

Managers need to ask themselves if their current big social data analytics platforms and capabilities are providing them with the necessary insights for improving their supply chains and brand building efforts.

Our analysis of the data from Twitter suggests companies are failing to respond to customers with the speed allowed by the platform.

While UPS responds in a matter of minutes – albeit usually to redirect to other channels – many others take hours and sometimes days to respond to customers.

Following up with a customer on the public platform after an issue has been resolved via DM or email would be a positive from a brand perspective. This is only happening in a limited number of cases.

Supply chain partners also need to better coordinate their Twitter-based customer service strategies so they are able to jointly resolve issues for customers and provide them with a seamless experience. Big social data can be harnessed to identify a range of process improvement opportunities but retailers and their supply chain partners need to be ready for change.

is a lecturer in Logistics and Supply Chain Management and  is a Research Fellow, both at the University of Sydney. This article was originally published on The Conversation on 6 April 2016.

Monthly Donation


Single Donation


Subscribe to IA for just $5.

Recent articles by The Conversation
La Niña could become the norm for Australia

Climate change is slowing down the conveyor belt of ocean currents that brings ...  
Big data: How we are being watched — all the time

Surveillance expert Ausma Bernot explains the many ways in which we are consta ...  
Reforming sexual consent laws faces formidable challenges

With sexual consent laws differing among states, a current push to harmonise these ...  
Join the conversation
comments powered by Disqus

Support IAIndependent Australia

Subscribe to IA and investigate Australia today.

Close Subscribe Donate