Hidden in the Business and Technology section of the today’s US print edition of the Wall Street Journal under Careers, there is a fascinating article, “Data Pushes Aside Chief Merchants“.
The article explains how continued and growing use of data, now big data and soon to be Internet-of-Things (IOT) created data (that will put big data into the shade), is creating ever new opportunities to automate merchandising decisions and push out the “gut driven” specialist that characterised the art of retailing.
The article is correct of course.
When you can embed the same decision making knowledge of the expert merchandiser into an algorithm, and you can feed the algorithm with as much data about consumer behaviour and preferences, the combination should lead to more productive and faster merchandising decisions.
The truth lays in the data, so to speak. The article gives examples for how this works.
I myself have been up close and personal with all manner of shopping basket and shopping mission analytics over the years as a supply chain planning, a researcher covering supply chain planning, and now covering the governance of the information that goes into those analytics.
Even though algorithms and big data/IOT are all the rage right now, the reality is that we had the same solutions all along.
We just have better versions of the technology, more data, and a greater willingness to experiment.
However, all is not well with the implied pending perfect science of retailing. The article explains how a science-only approach would fail.
If merchandising and supply chain decisions were based solely on data, failure would happen quite quickly. The example given was the innovative GoPro camera.
Current sales, demand and preference data, at the time of launch, was non-existent or showed lack-lustre interest from consumers.
As such, the science would say, “don’t merchandise any”. However, individuals knew that innovation of this kind would be a hit (a hunch) and so they were ordered, and success followed.
But even this examples if flawed. It assumed the “data” and the algorithm were all backward looking.
What if the algorithm was designed to weigh innovative new ideas for certain characteristics versus older products?
What if the data could be about connecting future relationships if innovations did sell?
The truth is the class of algorithms used today are quite old and not very forward looking. They are held hostage to data quality issues; they are not intuitive and relay on black-box capability.
So a new class of algorithm will be needed to balance pure science and some representative level of art. That is the future of retail.
Related to the article (but not mentioned in it ) is another challenge retailers face: what to call this new role that will blend art and science.
The Chief Merchant is the role that we are talking about - the right hand, number 2 role to the CEO.
Some retailers have hired a Chief Digital Officer, in part as a change agent, and also in part to focus on the outward facing customer channels.
But a digital focus on just one part of the business will only lead to some notional benefits; how will digital impact fulfilment, and customer service, for example?
For this reason we think that the role of Chief Digital Officer will be short lived, and will need to take a more enterprise view: we call this broader role Chief Data Officer.
We use that single name to represent all the variants and niche roles that are popping up from Chief Digital Officer to - let me see if you heard this one yet - Chief (data) Science Officer, and Chief Analytics Officer.
All such roles are just narrow segments of the ultimate role, Chief Data Officer. As we all become more and more digital, that will be the next long-standing board level role.
By Andrew White - Research Analyst, Gartner