Designing Digital Retail (Part 6): Building a Data Driven Business
This is the sixth in a series of articles that explore how traditional retailers can move through the challenges of digital transformation. You can see all the articles in this series here.
The next post will explore the value of design approaches for digital business.
McKinsey Global Institute (MGI) caught many people’s attention in 2011 when they estimated that retailers exploiting data analytics at scale across their organizations could increase their operating margins by more than 60 percent. However, while big tech companies such as Amazon and Google have capitalised on the opportunities and have built data-driven business models, most legacy organisations are struggling to realise the potential of big data for their bottom line. This is especially true for organisations with outdated legacy systems and closed architectures. Until these legacy systems are replaced and the business moves to open system architectures, there will be considerable areas of data darkness, which will disable the value creation potential of data.
So where do these businesses begin?
How can retailers going through digital transformation begin to create value from their data while they update their legacy systems and move to open architectures?
1. Discover and capture your data
Every second of every day retailers are generating massive amounts of ambient data. Transactions, customer support interactions, online browsing, stock management, inventory and replenishment, warehouse data, store fulfilment, online order fulfilment, supplier interactions and ordering, interactions with partners and suppliers, promotions and deals. These are just a tip of the iceberg of the ‘data exhaust’ of the business. A small percentage is being captured — but there are still vast quantities of bytes that are being lost. Some of this is structured data which can be captured relatively easily in traditional tabular databases — but there is also much that is unstructured which will need to be parsed and analysed to gain insights. The first activity of the business is to gather this data, capturing it in the cloud. Many retailers are establishing ‘data lakes’, and are beginning to set up pipes from their systems into this lake.
2. Aggregate and integrate your data
Retail organisations have many different sources of insight. Each of these by itself provides some useful information. However, these are currently separated, due to the lack of channel integration in the history of the business. Retailers need to get better at aggregating all sources of data. When these different data sources are brought together they provide a much richer and more valuable picture of the customer, and of the operations of the business. The organisation can gain insights into how the customer relates to the entire ecosystem of a business, rather than to a particular application or service. This single knowledge base can help drive decision-making, including both small changes to digital applications and wider strategic decisions.
3. Analyse your data
Having vast seas of data in the cloud is just the start. In order to draw meaningful insights, the business must therefore develop their data-analysis capabilities. Retailers must establish well-funded data analysis teams (also known as insight teams) that can begin to discover patterns in the data and relationships between factors.
There are immediately useful fundamental data-sets such as product and category performance, which can be matched against seasonal performance data or against promotions data. Clicks and sales on online applications can be easily tracked by each digital product teams, which can drive decision-making. Once these fundamental data-sets are being captured and tracked, the business can begin to run statistical analysis to search for other relationships. In reality most mature retailers are capturing these fundamental datasets. However, they often only exist in the form of Excel spreadsheets, and are their use is limited to small teams and divisions. The advantage of digital transformation is that these datasets can become available and usable across the business and can contribute to a deeper understanding of customers as they interact across the business, which can drive strategic decision-making.
4. Make sense of your data
It is not enough to simply see patterns in data. The business must also develop qualitative methods to interpret the data and understand the ‘meaning’ of the patterns. The best social science research uses a triangulation of mixed methods to understand complex phenomena. This must also be true for retail businesses as they seek to understand customers and operations. If the data suggests that customers are buying a product at a particular time of year, qualitative researchers can go out to find out underlying reasons driving that behaviour, so that the business decisions are made in light of the underlying need, rather than just responding to the surface phenomena. Furthermore, as retailers become more omnichannel, it is not enough to simply capture data about each channel. Organisations needs to develop research methods to understand cross-channel shopping behaviours, to gain insight into total customer experiences and behaviours. The most successful retailers will develop a capacity to care about and understand the customer. They will develop a capacity to observe them to see what they do, to listen to them to try to understand their thoughts and problems. This inquisitiveness will allow the business to find unmet needs or unmet opportunities, which can open opportunities for future growth.
5. Use your data to drive outcomes
The end question that all this points towards is, ‘How can this data drive outcomes?’ The simple answer is that whatever gets measured, gets done. If the business begins to capture data that can measure KPIs for operations, sales across physical stores, and digital products and services, they can begin to use that data to drive decision-making.
Retailers can here learn from Systems Theory, which has provided for us a clear description of the features of a successful systems in nature:
- Successful systems have a purpose: i.e. they are inherently goal-oriented.
- Successful systems have a means of measuring their progress towards their purpose.
- Successful systems have a means of measuring their internal state to ensure internal balance.
- Successful systems have a feedback loop in which there are able to modify behaviour in light of the measurements they make. This allows them to continuously move towards their purpose, and maintain internal stability.
- If a system is part of a larger system, the smaller system will have a purpose that contributes towards the goal of a larger system.
- A system is typically made up of multiple smaller systems and will therefore have to manage the internal balance of the smaller systems as it moves in a purposeful direction.
- Reality is fundamentally constituted from these nested, purposive systems.
Retailers must ensure these ‘features of successful systems’ are found at all parts of their organisation, at every level, within every team and in every process. This means alignment of purpose through the organisation, methods of measurement towards that purpose, and feedback loops to ensure the teams are moving in the right direction.
Ultimately the definition of success must be based in customer-focused measurements (such as ‘is the customer enabled to achieve outcome x?’ or ‘is the customer experience improved in this measure?’). This is a radical change from traditional management measurements, particularly at middle management, where definitions of success are typically related to hard accounting outcomes. Of course, hard accounting outcomes must also be measured. But the business itself should not be motivated by delivering customer outcomes, not by numbers. While this change may present a challenge to the business, retailers must understand that customer experience is the realm in which retailers are now competing. The challenge here is to link outcomes at middle management with customer outcomes. Unless a company can change the mechanisms of how the middle management of a company functions and is rewarded, design (and any other customer-centred approaches to business) will never be truly transformative.
Up Next: Designing Digital Retail (Part 7): The Value of Design Approaches in Digital Retail
