Imagine a retail store of the future. You walk into the store, the security cameras at the entrance capture your video, the system identifies you as a repeat customer, and retrieves your complete shopping behaviour and purchasing history. You get a custom welcome message on your phone, followed by recommendations on what to buy – essentials that you may have run out of at home, products that may interest you, and special offers on products you have previously purchased from the retailer. All this customised information reaches you and the store manager in real-time, activated by a simple video capture of your face, and by leveraging data analytics to deliver an exceptional customer experience.
Companies, in retail and other industries, are realising the potential of data analytics, and increasingly getting ready to adopt analytics to answer their key business questions. Becoming a data-driven organisation is among the ‘top 3 priorities’ for CEOs, found a recent KPMG study[i] on analytics. Hence, the adoption of data analytics should come top-down, with the objective to harness data to derive meaningful and actionable insights.
With data becoming a critical corporate asset (growing both in volume and complexity), organisations are increasingly capturing more and more of it – both structured and unstructured. However, the video of a customer walking in to a store, and logs of his/her purchasing history, are just pieces of information sitting on a server, unless leveraged to add value to the business and customer experience. And this is one of the biggest challenges: moving from capturing and storing data – ‘adoption’ – to analysing and utilising – ‘action’ – which is the key to deriving real value.
Data analytics, when done right, can lead to better decision-making, enablement of strategic initiatives, enhanced relationships between customers and partners, and improved ability to react to changes in the market. One approach we recommend, which delivers tangible and actionable outcomes, is to deploy analytics to a specific business question.
One of the UK’s top-four retailers wanted to understand the effectiveness of its in-store promotional initiatives and pricing changes. Although this is a common retail problem, companies tend to take the net increment in sales in a promotional period as the ROI, which leads to an inaccurate result. The solution developed by The Smart Cube included the creation of algorithms that isolated the impact of latent factors, such as pull forward, halo and cannibalisation of sales, allowing the true ROI to be determined. The outcome of ‘data analytics done right’ for the retailer has been significant cost savings (nearly £15m during FY2015–17). In addition, the solution has become a strategic tool: insights generated enable the business to formulate promotions down to a granular individual product level, and to devise a strategy that yields a positive ROI.
The retailer not only adopted and implemented analytics to derive insights from data; but it also applied those insights to its way of working. Yet, this next stage of ‘action’ is often a huge barrier in an organisation’s path to becoming data-driven. According to a research by McKinsey Global Institute, “Many (organisations) struggle to incorporate data-driven insights into day-to-day business processes.”[ii] One of the reasons behind this tussle with actionability is lack of trust in the data itself: KPMG found that 56% of CEOs are concerned about the integrity of data on which they base decisions; only 19% expressly stated they were not concerned.
Why does such a trust deficit exist? One answer lies in the quality of data and the capability of solutions – and a lack of both of them combined. For instance, if data is not collected with the right level of accuracy and rigor, any sample used for analysis will not be representative or may be inappropriately small. Equally, companies tend to have multiple sources of data, handled by multiple platforms and solutions, yet do not have the capabilities to draw them together to create a single comprehensive picture. In either case, the insights derived from analysis are likely to be inaccurate or incomplete.
The skills and technologies required for effective data analytics are varied, and evolving daily. For businesses to build these capabilities in-house can be difficult and costly, so many choose to use an external analytics provider – yet identifying the right one is also a challenge. There are four core attributes companies should consider in potential partners – 1) the solution offered (how sophisticated, granular and accurate the solutions are), 2) the way they work (how agile their way of working is), 3) scalability (not just providing a one-time algorithm, but creating a solution that can become a management tool and can be used by the client easily) and 4) cultural fit (an important parameter for companies while selecting a vendor).
Getting these factors right can be an invaluable step in progressing the journey from ‘adoption to action’, and can turn the video of a customer entering a store into a ‘wow’ customer experience.
[i] “Trusted Analytics Matter To CEOs” https://assets.kpmg.com/content/dam/kpmg/xx/pdf/2017/08/da-trusted-analytics-ceo-outlook.pdf
[ii] “The Age Of Analytics: Competing In A Data-Driven World” https://www.mckinsey.com/business-functions/mckinsey-analytics/our-insights/the-age-of-analytics-competing-in-a-data-driven-world