The retailer data journey

Modern retailing as we know it has been defined by periods of stability, interspersed with inflection points or – as they are fashionably known – disruptions.  Often these disruptions have created a technological or logistical sea-change involving data: for instance, the introduction of EPOS in the 1970s and 1980s, the rapid growth in the 1990s of loyalty programmes and retailer-owned payment cards and more recently, the e-commerce revolution.

All these examples have enabled retailers to effect change within their businesses to drive growth – whether this be through the development of category management brought about by systemisation of barcodes, or a deeper understanding of customer behaviour through analysis of loyalty card transactions. Using data to create insights that drive action should be the backbone for every retail strategy. But with the proliferation of data types generated today, organising your thinking to command the best value from your data investment is key.

From the early 2010s, the phrase ‘big data’ has come to be used to describe a whole set of trends and new concepts: wide data (lots of dimensions), long/tall data (lots of records), unstructured data (text streams, image etc) and fast data (real time or near real time). As there are simply so many new sources of data and new technologies generating data, it’s helpful for retailers to have a framework to understand this: A Shopping Trip Model.

A Shopping Trip Model, or customer journey map, should define the touchpoints that generate data and the context in which those touchpoints were experienced. In this way, any new data concepts can be understood as being a function of one of those touchpoints.  These touchpoints could include: need, notice, consider, visit, search, checkout, pay, obtain, use, share, reject (optionally, of course!) – it’s not a case of one size fits all.

For instance: a visit event is a touchpoint – a customer physically walks into a store, or accesses (and then perhaps logs in or is already logged into) a website, or opens an app – any physical or virtual space in which a transaction is made or influenced. What data is generated or could be generated by this event? When logging into a website, most retailers will now drop a cookie (which is primarily to allow for that visitor’s future web browsing to be understood and, perhaps, that visitor to be activated).

If the visit event was on a mobile device, the geo-location data can be captured: e.g. how close was that customer to a physical store or collection point when they visited the website? And beyond e-commerce, there is data to be gleaned in the physical store. Suitably enabled, a mobile device can indicate the time and point of entry in to the store (and geo-location can indicate where the customer came from). Increasingly video analytics technology is being leveraged to understand the age or gender of individuals entering the store; even using facial recognition in some cases.

So why collect all this data? At the heart of it, all retailing is about trying to create a better proposition for the customer: whether that’s cheaper pricing, a more relevant assortment, better staff service, more targeted offers or information. Aligning the core retailer brand proposition with investment in a data strategy is what great organisations do well.

We describe the retailer’s data journey by using a simple five step framework:

• Level 1: Essential Data – Any modern retailer should have and use this for financial and legal compliance: basic sales and HR data, for instance
• Level 2: Advanced Data – Retailers at this level typically capture a whole range of data which has latent value, but are only really using it for certain processes (e.g. RFID to manage stock control, loyalty card data to operate the loyalty programme)
• Level 3: Joined Data – This is where some retailers start to join the data meaningfully. It can be joined around customers, around stores, around time dimensions, around products, but requires some effort to put this data together for a specific end in mind
• Level 4: Enriched Data – These retailers continuously enrich data: creating segmentations that add descriptive dimensions such as ‘propensity to buy on promotion’ or ‘likelihood to have children’ and enable embedding enriched metrics into decision making
• Level 5: Data Partnerships – Retailers at this level recognise that there are data assets in the public domain that they don’t control, but can use to better understand and/or activate customers (e.g. social media data, wearables data)

An honest view as to where a company is on its data maturity journey and its alignment to the overall retailing strategy is essential to successfully moving forward. Many retailers are trapped by the highly operational nature of their business and find it hard to conduct this kind of review; data strategy, management and technology investment choices are not core capability for many retailers and this is why many have struggled to move forward at a pace that creates true competitive advantage. 

Jason Nathan is global managing director for data at dunnhumby.