In the world of retail data analytics it seems that the word insight is synonymous with the phrase "customer insight"; whenever I meet the Insight Team in a supermarket I discover that all analytics activity is focused on generating marketing-led customer insights.
Understanding customer (shopper) is, undoubtedly, critical to the success of every retailer but are customer segments, personas and voices the only artefacts of valuable insight?
Are the only factors that affect retail profitability divined from shopper spending patterns, demographics, polls, response rates, return rates etc? Might there be other insights on offer within retail data which get overlooked in a shopper-centric - no, shopper-exclusive - approach to insight?
Here are my Top Three other insight opportunities for those retail analysts, data scientists and Insight Teams brave enough to move beyond traditional shopper myopia...
1. Employee Insights
Customer engagement is critical to every retail operation - that's a given. It is not, however, the only engagement that matters; employee engagement is also critical. Until all customer-facing roles are replaced by artificial intelligence, customers are served by employees and shopping is a more pleasant experience when served by engaged, empowered and enthusiastic employees. It is no surprise that the most successful retailers at any point in time also have the most engaged, loyal and enthusiastic employee base.
What might effective employee insights look like? Smart retailers use employee data to look for the good and the bad - from outstanding 'upsellers' and brand advocates to potential patterns of till fraud and shrinkage. These are employee insights, even if they aren't labelled as such, and applying 'big data' analytics techniques to datasets which include employee markers will highlight a wealth of insightful patterns which help to replicate the very best behaviours and reduce or eliminate the worst.
Employee insights aren't only for shop-floor staff either - as digital dependencies increase it's appropriate to analyse GitHub activity, content engagement and customer service response rates across head office employees. There are even insights to be gleaned from interviewing staff and applying a little data science to the results - new product and service offers, employee satisfaction and hidden skills can be surfaced and harnessed.
2. Supply Chain Insights
If the modern retailer fears Amazon - and many do - then they should be looking beneath the surface of Amazon's much-lauded 1-Click™ customer experience to the extraordinary capability of its supply chain.
Arguably, supply chain optimisation represents the lowest-hanging fruit for many grocery retailers. Ocado aside, supermarket supply chains are still full of inefficiency through poor collaboration between retailer and supplier, based on limited information flow and even more limited analytics. I say arguably because I would argue this, working as I do in supply chain analytics (what we at Atheon describe as the "flow of goods"), but I remain staggered by the repetitive failings of large supermarkets and their suppliers to forecast promotional demand and then execute upon it.
Hundreds of millions of pounds are wasted every year promoting products which aren't physically available on the supermarket shelf - through poor planning and/or poor measurement during execution. Equally wasteful is repetitively shipping excess stock in the last few days of a promotion, leading to markdown and then waste as the stock reaches its sell-by date a week or two later.
"Data rich and insight poor" is a perfect description of the supermarket supply chain. At Atheon, through SKUtrak®, we are working hard to fight this but it can be a lonely mission whilst customer (retailer shopper, or FMCG consumer) remains the only insight worthy of consideration.
3. Product Insights
Surely product insights are really only customer insights? Perhaps, but exploring patterns of behaviour from the perspective of each category, subcategory and SKU can reveal a huge amount about shopper behaviour in aggregate and yet most (all?) product performance analysis is performed either at the (sub)category level or estate (not store) level.
The "category" team in most supermarket suppliers will examine product performance in aggregate - typically weekly, or even monthly, across entire retail estates. This made perfect sense 25 years ago when the syndicated (market) data providers began to collect and normalise EPOS data from supermarkets - this was only just feasible using the technology of the day. Nowadays, however, modern data processing, analysis and visualisation techniques allow for far more subtle insights. Perhaps some examples will help - how about these insights from our work with a UK-wide grocery retailer?
Cider, of course, sells particularly well in the South West of the UK - but whilst Somerset ciders sell consistently well across the whole of the South West, Cornish ciders only really sell in Cornwall. Oh, and spikes of sales for 2-litre Strongbow in Wales are more representative of University towns than a general love-affair with cheap cider by the Welsh.
Meat pies sell well in the North, but don't try to sell pies from Sunderland to Geordies, nor Newcastle-based brands to Mackems - a 'regional' approach to pie distribution in the North East resulted in simultaneously excessive waste and lost sales. This is a perfect example of where an average across the region suggests everything is fine - relatively low waste and high availability, on average - but store by store (and day by day) huge surfeits and deficits appeared.
A regional baker in the North West not only delivered both outstanding sales in the region - outselling Kingsmill, Warburtons and Hovis - but also generated excessive waste! How? Detailed store-level analysis highlighted demand all along the M6 motorway corridor - from Manchester down to Birmingham - but not further afield. So the regional stocking policy was stifling sales in North West Wales and the Black Country, but generating waste in Yorkshire and Staffordshire.
Yes - all of these Product Insights really demonstrate subtle shopper (and consumer) behaviours, but my point is that they were genuinely insightful and led to changes in ranging and distribution policy. The data necessary to uncover them had been available for years, but no-one had ever looked at it this way because the focus was on analysing shopper baskets for potential promotions and signs of loyalty.
Insight is defined (at dictionary.com) as..
an instance of apprehending the true nature of a thing, especially through intuitive understanding
... the word "customer" doesn't appear there at all.
I encourage all retailers - and any other businesses where the customer has become the solefocus of analytics and insight - to think more broadly, expand thinking and discover a world of profitable insights.
Guy started his career as a software engineer, and after progressing through technical and commercial roles founded Atheon Analytics in 2005, currently performing the CEO and CTO roles.