In January of this year, Amazon Go opened to the public in Seattle, U.S. In an increasingly cashless digital economy, Amazon Go pushed the boundary further with its cashier-less grocery store. While the jury is still out on whether this concept will be a game-changer for the retail industry, the fact remains that big data analytics, Artificial Intelligence (AI), Machine Learning (ML) and Natural Language Processing (NLP) came together to make this possible.
According to a WNS DecisionPointTM analysis, the retail sector is at a medium level of analytics maturity. The analysis is based on a survey of over 300 respondents consisting of chief marketing officers, brand managers, category managers, and heads of analytics and insights teams across industries such as retail, consumer packaged goods, insurance, banking and financial services, airlines and hotels.
While early adopters of technology in retail are taking the lead, this is the right time for the fence-sitters to start preparing their organization for big data analytics, AI, ML and NLP technologies. However, implementing these is a long-term strategy that requires time and effort. Here are some key areas where preparedness is required.
Building a Multi-source Data Strategy
Today, a vast amount of usable data is available on the customer. For example, in 2017, H&M teamed up with Google for the project ‘Coded Couture’ where an app collects data on a user’s lifestyle, travel routes, dining spots and daily schedules over a week to put together a ‘Data Dress’. The retailer was trying to provide a data-driven personalized solution to a customer’s problem of choosing the right attire for a day or occasion.
It is critical for retailers to not only identify the available data sources but decide which are the most relevant ones for their business as well. The wide range of data sources available include:
Data from website such as visitor traffic, clickstream, browsing pattern and cart abandonment
Data from in-stores such as customer interaction, aisle movement, reasons for returns and point of sale information
Data from mobile phones such as location and payment methods
Data from social media such as posts, tweets, images, videos, reviews and comments
One of the other challenges organizations face is accuracy of data. The analytics is as good as the quality of data that is given as an input to the system. Therefore, ensuring data authenticity is of significant importance. There are statistical methods available, such as t-test, f-test and regression analysis. These can help monitor the quality of data collated from various sources. Analytics platforms offer such features that allow for data checks at the input stage.
Organizational and Business Readiness
It’s imperative for organizations to answer questions around:
What will the data be used for?
What decisions will it drive?
Who will use the data?
Who will have access to the data?
Who will work on the data?
Retailers will have to define governance structures, accountability and the skillsets required. Readiness of the IT infrastructure to support the business goal is also necessary. As with any new technology, modern data architecture has its own challenges ranging from selecting the right tool stack, lack of large talent pool and integration with existing applications. Retailers who have taken a lead in this space are the ones who have been able to partner with niche product and technology players while staying focused on business outcomes.
Organizational readiness is a key challenge for retailers when it comes to leveraging the power of big data. Traditional retailers should start moving the needle on adopting new technologies by investing in big data and AI.
According to a Bloomberg report, AI-infused new technologies such as facial recognition and voice assistants are poised to cause the next big disruption in retail. With digital-native retail companies already way ahead of the curve, the time to look at analytics- and AI-driven platforms to process data and glean insights for traditional retailers is now.