And now, here is the first half of David's paper.
Reversal of Fortune –
From Crisis to “Supplier of the Year” in 18 Months
By David Poveda
Director Flowing Consultoría
Director, Demand Driven Institute, LATAM.
Last week, El Exito, the largest retailer of Colombia, awarded MIC “Best Supplier of the Year” in the category of textiles/clothing. MIC was selected amongst 500 companies, some of them extremely successful and that are known for their good practices in design, manufacturing and distribution. While that is a significant accomplishment in and of itself, what is even more significant is how far MIC came along in total performance in just 18 months.
MIC produces children’s garments under licenses from companies like Disney and Mattel. When you walk through their plant, you will see an amazing range of child favorites like Barbie, Mickey, Minnie, Hello Kitty, Spiderman and the like.
MIC was established in Medellín, Colombia in 1992 and has over 850 employees. Their products are sold through their own retail chain with 80+ shops in Colombia, Mexico and Venezuela. They are also suppliers of other retailers in Colombia and produce ladies garments for various direct sale companies. MIC designs the products, purchases the yarn, and outsources the fabric’s production and finishing. They have their own cutting and sewing plant to create end items.
My consulting company was called by MIC in December 2012. After the usual negotiation process, we started our implementation in February 2013. What we found there did not look good at all:
- Piles of inventory everywhere in the plant
- Unit cost reduction practices
- Everything purchased and produced to forecast
- Shipments to stores every two weeks in order to decrease transportation costs
- Silo mentalities across the business
- A central warehouse full of slow moving items
- Store managers very unhappy, etc, etc.
These are the typical consequences of the regular Push-and-Promote (which I call Push-and-Pray) mode of operations. MIC got to the point of having negative cash flow and they declared themselves in crisis. They asked for help.
I have been a consultant in operations since 1997, using TLS principles (TOC-Lean-Six Sigma). Some weeks before I started in MIC, I read a debate in the TLS Linkedin group by Chad Smith who was announcing some recently posted videos in a web site I never heard of. The video presented the 5 basic components of something called Demand Driven MRP (DDMRP).
New writings on MRP in the 21st Century?! They had to be kidding. It piqued my interest enough to watch the videos. I had no idea how much I was going to like the videos. I immediately bought an electronic copy of Orlicky´s Material Requirements Planning, 3rd edition, the book in which Carol Ptak and Chad Smith present a new blueprint for formal planning and supply order execution for the entire supply chain. In my opinion it set a new standard for this field, accounting for the emerging complexity and volatility that we see in today’s supply chains. I decided that this was something I should try with my new client. I went to Atlanta to take the Certified Demand Driven Planner (CDDP) Program. That course was taught by Chad Smith.
Fantastic stuff but I still had to go back and try to do my first implementation confronting a lot of obstacles. Maybe the most significant obstacle in MIC was the extremely long production lead times. Touch time for a regular production order is a few hours. The lead time/touch time ratio (one of the most important performance measures in operations) was around 200/1. There is no possible way that a company in the fashion industry (or any other in consumer products industry) achieves decent margins (if any) with such lead times.
Our first job was to substantially decrease production lead times. Work started. I spent time explaining Little´s Law. This is the fundamental equation of operations that too few people in operations are aware of. To me, working in operations without using Little´s Law is like asking a mechanical engineer to do her job without knowing that F=m*a.
Leveraging Little´s Law is what Drum-Buffer-Rope and Kanban do: keep WIP low and constant so the lead time is short and constant. Keeping WIP constant means that the amount of work released to the plant should be approximately equal to the amount of work that leaves the system.
I worked with the scheduling team to build a basic spreadsheet that would allow them to release orders in a synchronized manner. It was not a world-class solution but it worked fairly well. Lead time was reduced to 15 days in a few weeks.
At this point, we started our DDMRP implementation, going through its five components:
- Strategic inventory positioning: A dramatic improvement proposed by DDMRP. Systems Theory indicates that the best way of improving a system is by re-designing it, not by operating on it. Strategic positioning is the way in which DDMRP redesigns the system.
- Buffer sizing: A very robust mechanism for defining the size of the buffers and their zones. It is amazing how much information about a part you may get from its buffer once you get familiar with the way they are built.
- Dynamic adjustments: In order to size the buffer according to regular and expected market and product changes.
- Planning: Issuing replenishment orders with the innovative “Available Stock” equation that includes qualified demand.
- Execution: Following up open orders with high visibility and four different types of alerts.
Moving from the traditional Push model aimed to reduce unit cost, to a Pull based model that strives for protecting and increasing the speed of flow of materials and information is a significant and radical transformation for any company. The most difficult required change takes place in the mind of the management team. This is really the true achievement. The rest are details.
MIC´s implementation included the 5 components of DDMRP. We defined their supply chain starting in suppliers of yarns, fabrics and trims, went thought their production plant to a central warehouse and from there they shipped to all their stores and to clients in Colombia, Mexico and Venezuela. We sized the buffers and adjusted them dynamically as required by the market conditions and started issuing replenishment order to their stores, shipping daily to the shops in Medellin and twice a week every where else (daily shipment now is done to most of their stores in Colombia). The inventory at the retail level was reduced to the display requirements and they would only replenish what it was actually sold. The model was also implemented in El Exito, using VMI. Many people called them crazy but the company was ready for that.
The road was not easy and we had some difficult moments; the project was almost cancelled at one point in time. This shift requires a real champion with decision making power and influence. Fortunately Juan Esteban Gómez, Operations VP and one of the brothers who own the company, was (and still is) such a Champion. He managed to convince his older brother and sister to persevere. In doing so he probably saved the company.
MIC also understood the need for a capacity buffer in order to guarantee a fast response. They used to outsource around 40% of their production looking for cheaper labor/costs. When they embraced Flow, they decided to elevate their own sewing capacity (paying higher salaries) but actually being able to react within 24 hours to sales of high movers in the shops. I was informed a few days ago that MIC did not outsource any production during 2014. The higher salaries/costs in their plant are very much offset by the increase in availability of the right products and therefore, sales.