Lora Cecere, of the Altimeter Group, just published a great piece on the Supply Chain Shaman blog on why it’s time to take a leap of faith as far as using POS data to drive actionable replenishment is concerned. Yes, it’s time to move from PUSH to PULL using real demand and insight and not just seat-of-the-pants forecasts.
The post, which is over 2100 words in length, is too in-depth to do justice in a short summarization, so I won’t try, but I will point out a few keen observations that often go overlooked in most discussions and add one or two of my own.
First of all, as Lora keenly points out, you can’t prove ROI for an initiative before the initiative is done, which means if you want a proven ROI, you can’t be an innovator … and the big returns in this space will go to the innovators, not the renovators. You have to take that leap of faith.
There aren’t a lot of predictive analytic solutions on the market, but there really don’t need to be. If you get better data faster, you can re-run and correct your forecasts on a regular basis and minimize the the divergence between macro-level estimates and reality. That alone could save you millions.
You have to get close to the customer and get good at using the data in the sales relationship. If you don’t get close to the customer, learn their pain points, and figure out how you can use their data to help them, they’re not going to be that interested in helping you get access to it on a regular basis.
You have to build a cross-functional team led by business unit leaders focussed on innovation, or the initiative isn’t going to pick up enough momentum to make it. You have to break through mental barriers built up over years, or decades, of doing forecasting and inventory planning a certain way, and that’s not easy to overcome.
However, if you follow Lora’s advice, the rewards could be significant as stockpiles of obsolete inventory will quickly become a thing of the past. More importantly, and this is the one point the article should have really emphasized, so will costly long-term stock-outs. If you have access to daily POS data, you will not only see what is selling fast, and what’s not, but you’ll be able to run cluster analysis to see what products are selling well in what locales, and how the demand is spreading (outward or inward). This will not only allow you to quickly refill inventory on a popular, high-margin, item like a cellular phone or tablet PC, at a location about to run out, but predict which neighbouring locations should also be stocked up, and sense demand surges earlier in the cycle, giving you more time to ramp up production to prevent lost sales that could make or break the quarter.