This post has already been read 15320 times!
How long does it take you to drive to work?
You can probably give an answer that’s within a fairly tight range, say 15-20 minutes or 45 minutes to an hour.
One day, you leave your driveway at 6:30 and don’t arrive at the office until 9:00 because of a major traffic snarl. You’re accustomed to waking up at 6:00 and being at your desk by 7:00, but on this particular day, you left at 6:00 and were two hours late. So, you do what any reasonable person would do. When you get home that evening, you set your alarm for 4:00AM, right?
Why did that look just cross your face?
Because the idea of planning for something that is completely unpredictable is ludicrous and what has happened on any particular day isn’t necessarily a good indicator of what should happen on an average day.
Mike and I have been working in retail a long time and we see the same story play out time and time again.
- Replenishment VP looks at vendor on time fill rate report and he/she is not happy.
- Vendor inbound lead times are increased to ensure a higher level of on time delivery.
- Inventory holding costs go up a bit, but so do on time fill rates – for a time.
- Fill rates start to drift lower again.
- Rinse and repeat.
Taken at face value, none of this seems unreasonable. Many retailers even build systems that analyze historical order-to-delivery times and automatically update their lead times to stay on top of things.
The fallacy is introduced in Step 2 – increasing lead times may temporarily increase online fill rates, but this doesn’t necessarily mean that they didn’t have enough time to begin with.
I experience this trying to get my kids to bed. When I give them an hour’s notice to brush their teeth (a 3 minute task), they end up wasting 57 minutes.
No one will argue that some buffer and “up-rounding” is necessary when it comes to setting lead times in order to achieve some operational efficiencies, but you also need to make sure that the assumptions used in the planning process are consistent with the practices on the execution side. That principle is the reason that FedEx (whose entire business model is based on consistently meeting set lead times) is a $42 billion dollar a year business today.
For example, if you find that the trouble is with the vendor getting product ready for pickup on time, why, exactly is that the case? How long should it reasonably take to get an order ready for shipment if they had absolutely nothing else to do? If you sent a single order and started a stopwatch when it was received and stopped it when it was ready on the dock, how much time should have elapsed? A day? Maybe two?
So if you find that the vendor is consistently taking 5 or 6 days, where are the other 4 days going? Is it because they’re chronically out of stock when your orders arrive? Or because they are constantly getting surprised with large orders from other customers? You can add that 4 days to their lead time, but it doesn’t directly address the cause of the lateness.
With proper sharing of planning information over a long time horizon, the inventory should already be available when your order arrives. If they are getting “surprise” orders from their other customers, then those orders should be placed lower in the queue versus customers who are giving updated visibility to their future needs.
What if the vendor is getting shipments to the outbound dock on time, but the inbound transportation group is arriving late to the DC?Again, if you started a stopwatch when a truck leaves the vendor and stopped it when the doors opened at the DC, how much time should have elapsed? If it only takes 2 days to drive from source to destination, but shipments are consistently taking 4 days to arrive, what’s happening during those other 2 days? This can often be as a result of competing objectives within the retailer operation. Planning needs to have shipments on time to keep inventories low, but still maintain high outbound service to the stores. Transportation, however, is trying to optimize cube per load and will allow shipments to sit in a consolidation centre until sufficient volume exists to fill a truck (regardless of the due dates on the orders).
In this scenario, there needs to be a bit of give and take on both sides. On the one hand, Planning can’t expect transportation to ship a single pallet point to point across the country. On the other hand, transportation can’t hold shipments indefinitely so that they can pack trailers to the brim before moving them. In this case, it’s entirely appropriate to add consolidation time into the transit time to achieve consolidation efficiencies, but once the consolidation time has been agreed to, then due dates must be strictly enforced. If some trucks have to go 75% full instead of 90%, then so be it.
Next time you’re reviewing your lead times, pick a couple of problem lanes and try digging a little deeper and see what can be done to fix operational issues that are causing lateness.
Adding more buffer to lead times should be the last resort, not the first line of defense.
More Stuff
Click here for a simple, straightforward article on lead time reduction from Logistics Quarterly by Robert Martichenko. A short, easy read.
Photo by LifeSupercharger
- Small Data, Big Insights - July 14, 2016
- Is Flowcasting the Supply Chain Only for the Few? - June 3, 2016
- Philosophy - February 12, 2016