Picture this in a luxury retail store while one shopper is browsing and another, sitting on a sofa, browsing their web store via their smartphone and comparing the product makes and prices. This is the retail in digital world where shoppers expect consistent experience across various channels of the brand.
Beyond the bricks-versus-clicks divide, leading retailers like Walmart, Costco, Target to name a few adopters of big data embodies the contrary forces that scientists are harnessing by mining digital information to shape pricing decisions. These moves to unlock the power of big data, are increasingly redrawing the relationship between retailers and consumers.
The retail industry traditionally has been generating detailed data on consumer behaviour and purchase history through various transactional and customer relationship systems. Exploiting these data to increase basket value and optimize margin remains a challenge with traditional technologies of business intelligence tools and data warehouse architectures. As consumers are empowered by data and mobility, with access to competitive prices and information while visiting stores, retailers capacity to integrate data from online (browsing pattern, clickstreams, social media) and in-store customer experience (POS, surveys, CRM etc) and be able to react in real time is key.
Business Drivers in Retail
Deliver a consistent, personalised product mix and pricing to customers across all channels
Offer a differentiating customer experience
Obtain a deep understanding of customer behaviour to increase loyalty
Increase conversion rate and average basket value due to efficient and targeted marketing campaigns
Reduce cost by optimising inventory, planogram and supply chain management.
1. Next Best and Personalised Offers- Recommendation Engine
Big Data enables to better profile the customer and use collaborative and context-based filtering to offer the most appropriate product, product bundle, or offer at any given time.
Using Big Data technologies to store all customer data, including all customer interactions, and progressively adding social media and social network analysis will help retailers optimise customer experience on all channels and improve sentiment towards the brand. For retailers marketing teams, it means finally being able to engage their end consumer at a more personal level, effectively pushing the right product or promotion recommendation at the right time. When the customer gives permission, the machine learning application uses recent spending histories and customers profile data for a huge number of transactions to train a recommendation model. The model predicts which products and offers a particular customer might enjoy and makes those recommendations through different channels- mobile or in-store executive.
2. Store Design and ergonomics- Optimised Planogram for effective Customer Engagement
Big Data techniques allow the capturing of geo-spatial data to drive insight into customer's movements within stores. This enables the mapping of hot and cold spots. With this insight, a retailer can improve floor layouts by enticing customers into cold spots or moving high revenue generating products into hot spots.
Retailers can use this space and floor planning information to drive more impactful, more efficient category management. Data around floor plans can also be combined with information on past sales behaviors (shopping history, online activity, etc) to create tailored navigation path and shopping lists for customers, which lead them through a store in the most efficient manner. In-store navigation technologies enable retailers to build mobile applications that significantly improve the in-store customer experience. Retailers could do this by offering customers the most optimal way to purchase items on their shopping list, and highlighting appropriate complementary products or promotions along the way.
3. Real time Inventory- Data-Driven Stock and Ordering
Big Data solution can help retailers optimise supply chains to reduce cost, improve service, and gain vital insight. Big Data analytics are used to predict inventory positions in stores and distribution channels. This is achieved by utilizing demand plans and forecasts, sales history, external predictors of future performance such as category trends, weather patterns, local events and so on, enabling retailers to reduce both out-of-stocks and over-stocks.
Big Data analytics can also deliver full supply chain visibility by capturing real-time inventory positions across the enterprise and through the extended supply chain. Data to be leveraged include open purchase orders, in-transit inventory, or vendor and distributor inventory. This information is critical for retailers looking to deliver an omni-channel shopping experience to their customers. .
4. Assortment Optimisation
Deciding which products to carry in which stores "assortment optimization" based on local demographics, buyer perception, and other big data inputs can increase sales materially. One leading retailer I worked with, for example, used consumer research, market and competitive analysis, and detailed economic modelling to identify the causes of its flat and declining growth at the category level. For the brick-and-mortar retailer, substantial gains can be found by optimizing the placement of goods and visual designs, such as end caps and shelf placement, by mining sales data at the SKU level. In essence, a traditional retailer can build a more granular, localized version of the kinds of optimization that take advantage of aggregated foot-traffic data. For online retailers, merchandisers can adjust web site placements based on data on page interaction such as scrolling, clicks, and mouse-overs. Tools like Apache Spark can enable both SQL-based and more advanced machine learning models to be run against the large and diverse data sets.
5. Real time Pricing Optimisation
Retailers today can capitalize on the granularity of data and the move to a finer, more relevant consumer engagement model for pricing and use higher levels of analytical horsepower to take pricing optimization to a new level. Business analysts can employ complex demand-elasticity models that examine historical sales data to derive insights into pricing at the SKU level, including markdown pricing and scheduling, and even correlate product pricing changes to more detailed segments, if not to the individual consumer. Using this approach, retailers can begin to realize personal pricing schemes powered by multi-channel engagement models. Retailers can also use the resulting data to analyze promotion events, evaluate sources of sales lift, and interrogate underlying costs that these activities might entail. These sophisticated models can be created with machine learning in Spark and executed natively in Apache Hadoop.
Bricks-and-mortar retailers have for years been grappling with the reality that increasing numbers of consumers are buying everything from apparel and books to designer furniture online, and many have yet to devise a winning strategy to integrate their physical and e-commerce stores.Now is the right time for retailers to start their big data initiative. There are many different use cases and substantial benefits to be attained from analysing Big Data.
Retailers who have implemented these Big Data Use Cases have gained insights that are driving innovation across merchandising, marketing, sales, inventory, and supply chain processes to generate more than millions of dollars in incremental profits. For example, text mining and sentiment analysis are helping staff identify trends across digital channels. This is helping enable customer-driven trend decisions and drive multiple increase in the sale of store-branded items. Likewise, with more comprehensive market basket analysis related to promotional items, the retailers have achieved a substantial increase in sales volume and revenues.
Another exemplary quote from Sam Walton, Walmart's founder "High expectations are the key to everything". Big Data technologies powered by Apache Spark and Machine Learning capabilities can help retailers meet high customer expectations in this modern digital world.