Over the years, working for and with numerous manufacturing companies, I’ve seen many supply chain practices that cost companies money. Over the next several weeks, I’ll outline these issues and discuss some ideas around how to avoid these practices. You can find the previous posts here:
Making decisions based on bad data (supply chain data accuracy)
I went into a store the other day. I’d driven an hour to get there. I went to that particular store because their web site confirmed that they had 12 units on hand of the thing I was looking for. When I got there and went looking, I couldn’t find it… the slot was there but there was nothing on the shelves. I found someone from the store and asked about my item. Yes, the computer shows they had 12 units on hand. They went looking at the shelf the computer said the item was on (the one I just checked). Not there. They looked in the back. Nope, nothing. They searched the shelves around where the item was supposed to be. Nowhere to be found. “I’m sorry sir. It looks like the computer made a mistake…we don’t have any”. Hmmm… So I went back home and ordered it from Amazon.
A couple things struck me about that interaction. Having wasted time going to that store, I’d be less inclined to use that store in the future – at least I’d be much less likely to trust their website’s inventory. The second is that it likely wasn’t the computer that made a mistake, it was a mistake made by a person or process somewhere along the way. And finally the same types of mistakes and process failings that resulted in my wasted trip occur all the time in supply chain. In addition to losing customers like me, those mistakes result in bad data that cost manufacturing companies millions of dollars.
Bad data in supply chain seems like it should be a minor thing. I mean, it’s just numbers right? Let’s look at some typical supply chain data errors and think about the potential costs;
If the planned lead-time is longer than actual, you get excess inventories. If the planned lead-time is shorter than actual, you get stock-outs and late customer orders.
Bill of Material
The Bill of Material drives material requirements through-out the supply chain. Missing components, extra components, bad effectivity dates, incorrect quantity per values will result in excess inventories, scrapped items, inaccurate costing, late customer orders, stock-outs and increased WIP.
Cost and price data
Inaccurate price and cost information impacts decisions based on margin. It can make unprofitable products appear to be profitable, and profitable products appear unprofitable. In accurate costs can also impact pricing decisions driving higher or lower prices based on cost assumptions. Finally, inaccurate cost information may cause you to make incorrect sourcing decisions.
Part master – ordering rules (Lot sizing, policies, etc)
Part master data controls how the system creates new supplies. Bad data here can cause you to order too much driving excess inventory or too little driving additional ordering costs.
On Hand Inventory quantity/status
Decisions on when and how much to order are based on how much inventory the system thinks you have. If inventory quantities or status is incorrect, excess inventory, excess costs, late customer orders and stock-outs can be the result.
The routing table describes how material flows through the shop and includes information on how long work should be scheduled across each work-center. If this information is wrong, material will end up at the wrong work-center or the right work-center at the wrong time. As a result, you can expect increased WIP, late customer orders and stock-outs.
Safety Stock rules/quantities
Safety stock influences the inventory levels expected on an item by item basis. In addition to mistakes when calculating safety stock or setting safety stock rules, safety stock values can get stale. In other words, the inputs (demand variability, lead time, etc.) that went into setting the safety stock value at the time they were set, may not be valid now. Inaccurate safety stock values can drive excess inventory, excess costs due to expediting, late customer orders and stock-outs.
Demand (Actual Orders)
Inaccurate order information impacts in two ways; 1) an unhappy customer – especially if they don’t get what they ordered and 2) Increased costs due to excess inventory, excess costs from expediting.
Forecasts are always wrong. Yet forecasts are what drive the business in many industries. When forecasts are wrong, the result is too much inventory of some items and too little inventory of others. This puts you in the unenviable position of having to explain to your stockholders why you have excess inventory while at the same time can’t meet revenue numbers because of stockouts.
Historical demand is the record of what was sold and when. Historical demand is used to drive statistical forecasts, safety stock calculations and more. Errors here will cause inaccurate forecasts, excess inventory, excess costs, late customer orders and stock-outs.
Capacity information determines what amount of work can be done in a given work center. In finite systems, orders will be moved around to respect the available capacity. In infinite capacity systems, planners will manually move orders around to level load work centers. If capacity information is wrong, you can expect to see overloaded or under-loaded work-centers, excess costs due to expediting, increased WIP inventory, late customer orders and stock-outs.
As you can see, there are a variety of ways that inaccurate supply chain data can cost you money (and I’ve really only scratched the surface here). So what can be done? There are a number of tactics that can be used to improve supply chain data accuracy;
- Audits – At a company I worked at in a previous life they recognized, through an earlier failed ERP implementation, the importance of data accuracy. As such, they implemented a weekly MRP data audit where several parts were picked at random and the part master, BOM and Routing parameters were checked and validated. In my previous post, Reason #6 Not effectively managing inventory, I talked about cycle counting. This is a similar type of auditing targeting inventory accuracy. In both cases, the key to improvement is root cause analysis. If data is wrong, simply correcting the problem just fixes the data issue for that part. Understanding how the error occurred and fixing the process that caused that error means that the error is less likely to happen again.
- Data responsibility/data security – Ensuring that the right people are responsible for a given segment of data and that only those people can change that data is difficult and can be frustrating for some. However, the risk of not locking down the data is that there is no control… anyone can change anything. In the vast majority of cases, people won’t sabotage data (although that can happen). No, usually, bad data gets created because people make changes without fully understanding the impact of what they have done. Again, in a past life, we had one lady responsible for all item master changes made to the ERP system. Any change needed to be requested via a paper form. This form needed to have various approval levels before she would make the change. While it did slow things down and admittedly was relatively inefficient, it ensured that any item master change was well thought out and vetted before it was made. As a result our data accuracy percentage was consistently in the high 90s.
- Alerting and what-if – With more advanced planning tools, you have a few more options around how to identify and correct some types of data errors. Automatic detection of errors like missing cost data, missing sourcing data, order policy mismatches can all be detected and alerts generated that inform those responsible of potential issues. In addition, advanced planning tools enable, through the use of simulations, the ability to try different planning parameters to see what the impact would be without actually driving any change to actual production.
Data errors if uncaught can result in millions of dollars of losses. That being said, with some focused effort and perseverance you can eliminate the majority of data errors and get your supply chain running like the well-oiled machine it should be.
What data errors have you seen and how did it impact your supply chain? How do you manage data errors in your supply chain? Comment back and let us know.