Data Rich, Information Poor: The Conundrum of the Distribution Center
We live in a world that is bathed in data...and yet too often that data yields little in terms of information. Today, data proliferates business, but the worth of that data can only be measured by the insights derived from it.
Further, we need data that we can substantiate from other relevant metrics and that we can act upon effectively. Currently, we lack (and hopefully are evolving) a framework and a systematic approach to analyzing even unstructured data in an unstructured manner. We need to track the intelligence gleaned from it until we find interesting insights that impact a real business problem.
That is the next frontier of data analytics, which can be further substantiated by simulations and or even prediction markets. Consider this scenario: A corporate project is potentially in trouble. Everyone in the organization knows it, except the vice president who owns the project. The failure comes, and the company brand is soiled and the VP and his people are all fired, and the company's stock price tanks. What if the employees, those on the ground, were allowed to share their belief on the potential outcome in a regulated way? Then, the VP would have data around the potential perceived success. Of course, this wouldn't work in every scenario (which as when the management team is trying to effect revolutionary change), but it's interesting to think about.
During a post pilot review meetings in which I was involved, a distribution center (DC) manager expressed his frustration about the three to four hours daily that his supervisors were spending collecting data on the organization's operators in order to answer seemingly simple questions like:
I remember one client that was plagued by a warehouse management system (WMS) that, for various reasons, wasn't producing the expected throughput. I went to a senior manager but was met with nothing but anger and frustration. He was angry ,because the DC throughput was almost half of what was promised to them during the WMS sales cycle. Eventually, I figured out the bottlenecks that were slowing them down, addressed them and helped them get back on track. At the end of the project, here are the questions that were looming:
With the evolution of Internet of Things (IoT), data is proliferating and mobile devices are loaded with all kinds of sensors, accelerometers, gyroscopes, pedometers etc. It is often possible to gather vital intelligence to measure the efficacy of the processes and operations in a distribution center pretty quickly.
In any DC, the biggest expenses are labor and supplies. Often, DC personnel don't focus as much on inventory cost because it is outside their control. The inventory/merchandise needs to be really managed throughout the whole supply chain. If it is something that is owned by the retailer or distributor, then they care about the accuracy of the inventory count and location of inventory.
As I said, the two most important expenses executives care about are the labor expenses and supplies. While managing labor, it is very important to manage unproductive travel time, for example. Unproductive travel time can be measured easily by monitoring the number of steps the operators walk or travel during the entire course of the day. If an operator that walks a lot but gets little work done, it needs to be flagged and addressed. That's just one way that data could be actionable.
Let us know what activities you monitor and how you leverage that good information to optimize spend in the comments section below.
Further, we need data that we can substantiate from other relevant metrics and that we can act upon effectively. Currently, we lack (and hopefully are evolving) a framework and a systematic approach to analyzing even unstructured data in an unstructured manner. We need to track the intelligence gleaned from it until we find interesting insights that impact a real business problem.
That is the next frontier of data analytics, which can be further substantiated by simulations and or even prediction markets. Consider this scenario: A corporate project is potentially in trouble. Everyone in the organization knows it, except the vice president who owns the project. The failure comes, and the company brand is soiled and the VP and his people are all fired, and the company's stock price tanks. What if the employees, those on the ground, were allowed to share their belief on the potential outcome in a regulated way? Then, the VP would have data around the potential perceived success. Of course, this wouldn't work in every scenario (which as when the management team is trying to effect revolutionary change), but it's interesting to think about.
During a post pilot review meetings in which I was involved, a distribution center (DC) manager expressed his frustration about the three to four hours daily that his supervisors were spending collecting data on the organization's operators in order to answer seemingly simple questions like:
- How much work was done? How many picks were done?
- Did the operators meet their targets?
- Did anything hinder the operators from doing what they were supposed to do? (For example, with replenishments, often the pickers will have to wait when the bins run out of product because the bins will have to be replenished from the reserve locations.)
- How are new hires performing compared to the veteran workers?
- How long does it really take to fully onboard the new hires?
- How long does it take for new hires to perform consistently and match the corporate standards?
I remember one client that was plagued by a warehouse management system (WMS) that, for various reasons, wasn't producing the expected throughput. I went to a senior manager but was met with nothing but anger and frustration. He was angry ,because the DC throughput was almost half of what was promised to them during the WMS sales cycle. Eventually, I figured out the bottlenecks that were slowing them down, addressed them and helped them get back on track. At the end of the project, here are the questions that were looming:
- Hey how do we make sure we keep this going?
- How do we make sure we maintain this level of efficiency and productivity?
- How do we make sure we all know that we are consistently performing to these metrics? Or there is an alert or a red flag that pops up if we slack?
- Also how do we keep ourselves on top of this, especially when the business is changing dynamically? A good example is shift of commerce from Retail to e-commerce and to mobile these days.)
With the evolution of Internet of Things (IoT), data is proliferating and mobile devices are loaded with all kinds of sensors, accelerometers, gyroscopes, pedometers etc. It is often possible to gather vital intelligence to measure the efficacy of the processes and operations in a distribution center pretty quickly.
In any DC, the biggest expenses are labor and supplies. Often, DC personnel don't focus as much on inventory cost because it is outside their control. The inventory/merchandise needs to be really managed throughout the whole supply chain. If it is something that is owned by the retailer or distributor, then they care about the accuracy of the inventory count and location of inventory.
As I said, the two most important expenses executives care about are the labor expenses and supplies. While managing labor, it is very important to manage unproductive travel time, for example. Unproductive travel time can be measured easily by monitoring the number of steps the operators walk or travel during the entire course of the day. If an operator that walks a lot but gets little work done, it needs to be flagged and addressed. That's just one way that data could be actionable.
Let us know what activities you monitor and how you leverage that good information to optimize spend in the comments section below.
No comments:
Post a Comment