In today’s connected world, shaping the customer journey has become increasingly complex. People interact and collaborate with each other using a combination of e-commerce, social media, and mobile. In this new ecosystem, people can easily access information such as product and brand reviews, which gives retailers extra marketing power but also makes them more vulnerable to customer criticism.
The modern customer has a large and dynamic digital footprint, data-driven analytics across all channels can yield many insights, which can help retailers better engage with their customers. But, as commented by Jeff Rosenfeld, the VP of Customer Insight and Analytics at Neiman Marcus: “Too often people think simply collecting data will lead to insights, but the only way it will have an impact is if people start using it.” So, three steps must be taken by retailers: collect the data, analyse it, and most importantly – act on the results of your analysis.
Assuming that you already have a customer signup and data capture processes in place (e.g. via a loyalty scheme), let’s have a closer look at analysing customer data and how to utilise your findings. From my observations of the retail market, the number of retailers focused on collecting customer data keeps growing, however, few have managed to effectively use this data to grow their businesses.
We can classify analytical processes into three types:
Retailers need to master all three in order to remain relevant in the modern, and continually evolving, economic system.
About 2.5 quintillion (2,500,000,000,000,000,000) bytes of data are being produced by customers and businesses every day. The data is generated by practically every internet-connected, digital event, e.g. when customers click on an advertisement, browse or like a product, add an item to cart, checkout and make a purchase. Retailers also collect customer data via social media, digital advertisements, and marketing. The question arises: with such an ocean of information, why have so few retailers actually managed to pipe into it?
The answer surprised me: it seems to be self-delusion. According to a survey conducted by Bain & Company, more than 80% retailers believe they are driving an exceptional customer journey, but only 8% of their customers express satisfaction with the experience.
So, the first step in your quest to improve your customers’ journey must be a sober assessment of your current use of the available customer data. I have outlined below what retailers need to do if they want to master customer-related data, so they can truly engage and collaborate with their customers in a meaningful way.
Level one, the simplest layer, means the elimination of customer data fragmentation and duplication within the retail enterprise. All customer-related data must be stored in a single system: customer details, addresses, contacts, purchase history, as well as loyalty arrangements. Even this basic model eludes many retailers, as they have failed to put in place centralised management for the data flowing in from different channels.
Once the customer data has been unified, the next level can be achieved – customer data enrichment by automated interactions with social media. In today’s tech world, customers commonly use several touch points for a single sale, for example: inspiration from Pinterest, product browsing using mobile, purchase via desktop website, return in the nearest brick and mortar store, and an opinion left on a consumer blog.
Therefore, to truly understand the customer journey, data from all these multiple channels needs to be captured. This represents a challenge and it requires website scraping, with continuing refinement of the scraping algorithms, as not many social channels provide structured data access. Note, that as this data starts flowing in, the structure of the retailer’s customer database must be expanded, to accommodate the enriched information.
Once a retailer’s customer management system gains the ability to capture and store both internally and externally generated information, data segmentation and analysis can start in earnest. At this stage, at level three, retailers can generate personalised, rich customer offers and enter into individualised (but automated) correspondence. Most retailers would be content to have such abilities.
However, there is also one level up – the fourth layer. It allows retailers to move beyond business intelligence (BI) and enterprise data warehouse (EDWs) software, because it involves dynamic database structure evolution, based on the nature of the collected data.
At this level, the system automatically detects and recommends new data entities for creation, which once approved, can be used by the retailer’s system to start generating relationship data. For example, if the system detects that the retailer’s customers visit a certain shopping centre more frequently than other centres, it may recommend that such centres are recorded in the database so future trips can be tracked through explicit relationships. I am reluctant to use the term ‘Artificial Intelligence’ here, but fundamentally level four means a self-learning system.
In the above context, retailers have two immediate challenges:
Once this analyse > learn > act model is in place, the quality and quantity of customer-related data can be continually expanded in the background, making customer communications increasingly more effective.
This base analytical system requires a centralised database to integrate data arriving from multiple channels, so the retailer must decide on, implement and maintain such a database. As most customer analysis routines require access to sales information – what was bought, where and when, but even more importantly, what was not bought – in most cases the choice is simple: the database that stores the majority of customer transactions must be the master, to minimise data movements across the retailer’s systems.
With a single view of all data in place, the customer journey can be monitored by getting people to register as a customer and then recording viewed items, time spent on specific pages, items in the basket, items in the wish list, items left in suspended sale transactions at POS, returns, verbal / email interactions, ratings etc.
This internal data can then be enriched by integrating core retail systems with social media platforms like Facebook, Instagram, and Twitter. Using site scraping algorithms to connect customers’ social conversations with real-time order information gives retailers a much richer and comprehensive view of their customers’ buying journeys.
The unstructured insights gathered from social should be combined within the same view as the real-time customer order, transaction history, and tendency data to enable the automatic exposure of possible issues in the buying journey within the context of a customer’s history with the retailer. Issues can then be corrected before they become problems, and the experience can be continually optimised and tested.
Once a retailer learns how to effectively and holistically analyse what motivates their customers to buy, essentially the creation of 360-degree customer profiles, these learnings can then be translated into frictionless personalised promotions and shopping experiences across all channels.
The retailers who use data to make their customers feel like a “market of one” and back it up with great products and a seamless shopping journey will grow into giants on the world stage.