As supply chain planners, much of what we do is to anticipate future outcomes and plan our activities around those expectations so as to optimize business results. What do you do though in the aftermath of a year like 2016 that seems to have upended so many conventional wisdoms and when so many experts so badly missed key outcomes and developments?
Nobody saw it coming...
2016 saw analysts and experts fail in reading polling data – the US presidential election or the UK’s vote to leave the EU. Or for that matter, analyzing economic trends – the Dow at 20,000 or the Fed’s dovishness for most of the year was contrary to most Wall Street’s predictions at the start of the year. Experts at Samsung failed to see the disaster that would be their inaction on the Galaxy 7 phone and I (a passionate sneaker fanatic) still cannot understand how a simple Snapchat by two Californian high schoolers could drive the price of a pair of Vans shoes to $400,000. Damn, Daniel indeed.
So is 2017 the year supply chain managers start to putting a bigly safety margin insurance on demand planning models?
Or do you double down? Re-invest to bring in much needed innovation and technology into your demand and supply chain planning process to build a more robust, ground up, integrated and dynamic planning environment that increases inventory turns, service levels and other supply chain KPIs?
As we enter into 2017, now more than ever it is critical for supply chain managers to strengthen their supply chain planning and analytics capabilities.
While I am personally hopeful that trade wars are more rhetoric than actual policy actions – 2017 could see many retail, CPG, or consumer electronics supply chain managers having to re-draw and re-optimize their supply chain networks. Now more than ever will supply chain managers have to pay attention to how their strategic network design choices affect their flow through tactical and operational supply chain results.
Second, old assumptions = missed forecasts. One interesting post-election evaluation of polling data I read talked about how a heavy reliance on phone sampling mis-lead pollsters as phone sampling no longer made sense in 2016. Similarly, I recall when onboarding several of our new retailers or wholesaler clients, we oftentimes discover over decades old assumptions about safety stock or days forward coverage factors in their supply chain planning spreadsheets even though lead times had dropped by over 50% in the same period.
Dependence on worn-out assumptions are incredibly hard to spot and fix when generating statistical demand forecasts at an aggregate level which then gets disaggregated down to SKU or location levels. With rising volatility in business environments – this 90’s approach to demand planning will most certainly fail in 2017. Instead, supply chain managers need to look for advanced machine learning algorithms that constantly tests and recalibrates key assumptions based on the latest data and social media signals to generate detailed ground up statistical forecasts that provide more rigorous inputs for your supply chain plans. If 2016’s forecasting debacles teaches us anything, it is not that data science is dead but that the set it and forget it “black box” approach to demand and supply chain planning is no longer tenable.
2017 will be a year of change. As we prepare ourselves to welcome in the new year - I wish all of you a very happy holiday season and New Year from the Solvoyo family. May all your plans come to fruition and you have a prosperous 2017.