Predicting the Unpredictable
Jim Correll, director Fab Lab ICC at Independence Community College, Independence Kansas
Many people don’t know that I did two years of hard time at the Amazon fulfillment center in Coffeyville between 2001 and 2003. At the time, the now defunct facility served as “overflow” regulator, serving all parts of the country. Facilities on the east and west coasts fulfilled orders roughly based on the customers’ geographic locations but when there was an overflow of orders at any other facility, the orders came to us. This made for a wildly fluctuating, seemingly unpredictable workflow, literally a feast or famine of shipping and receiving to do.
Among the most unpredictable workflows was the process of receiving individual items where less than case lots were needed. The individual receive line—we called it Rambo receive—consisted of one team unpacking boxes full or random items and scanning them into the system. Another team took totes of the random items to an area called random stow where they were put on shelves anywhere there was space. About every twelve inches of shelf was labeled with a unique location number and bar code. As the stow-worker put an item on the shelf, they would scan the item and then scan the location label so the system would know where the item was. There was a total of nearly 400,000 locations in the facility and when “pickers” went on a route to pull items to fill an order, the system would tell them, on their hand-scanner screen, the location where they would find each item.
The work flow for Rambo receive fluctuated as wildly as any other in the facility and the task of having the right size crew for the work each day fell upon the Rambo area manager. That was me. I had about twelve full-time employs on the Rambo crew yet some days the work required up to 45 people. At that time, Amazon came up with a concept called the “flexy” worker; part-time folks who agreed to be “on call” with one day’s notice to come in and work when we needed them. The Rambo manager was the one that made the call on a day to day basis of how many people to bring in the following day.
The incoming Rambo product was unpredictable. Amazon provided forecasts which were so inaccurate as to be nearly useless. It was a no-win situation for the Rambo manager, i.e. me. Backlogs of items to be received prevented timely filling of customer orders, a definite no-no. Bringing in too many people resulted in low productivity numbers and that was unacceptable too. So, the challenge was to take an unpredictable situation and make it predictable. How was I supposed to do that?
After the first couple of months I actually did get pretty good at having the right sized crew on hand each day, but it wasn’t some scientific formula based on forecasts—remember the forecasts were no good. I was not able to find a working crystal ball, so I didn’t have that advantage either. I can’t really explain how or why I became good at staffing but here is what I think. I was a “working” manager, meaning I helped do the actual work when I could. At first this was so I could learn the processes, but later I continued helping with actual work to show the team I was not too good to do the work. I believe that exposure to the work helped me develop a 6th sense or “pulse” of the process and somehow that 6th sense gave me a “gut” feeling of how many team members I needed to bring in on a daily basis.
A world where business and government leaders constantly harp on the need for “data driven” decisions doesn’t make allowance for decision making that is required when there is insufficient data. This leaves room for trusting your “gut” and I know many successful entrepreneurs who have come out of the closet and said they rely on “gut” decisions and predictions. We can all develop a 6th sense, that improves over time, of predicting the unpredictable. This includes utilities and contractors who throw up their hands in trying to predict what hour or even day they will be able to show up at your site. While the 6th sense does not always lead to perfect decisions, in most cases its better than the consequences of seemingly unpredictable processes for which we throw up our hands and give up managing at all.
Jim Correll is the director of Fab Lab ICC at the Center for Innovation and Entrepreneurship on the campus of Independence Community College. He can be reached at (620) 252-5349, by email at firstname.lastname@example.org or Twitter @jimcorrellks.