Executive Briefings

Traditional Supply-chain Solutions Don't Work for Contract Manufacturing Model

A conversation with Jeff Range, vice president of operations at net.com, Fremont, Calif.

Network Equipment Technologies Inc., doing business as net.com, develops and delivers platforms for broadband, internet protocol telephony and multiservice networks. An architect of the networking industry, net.com has been supplying service providers, governments and enterprises around the world with reliable networking technology for more than 20 years. Jeff Range, a native of the U.K., joined net.com in 2002. He began his career as an engineer with Marconi Space and Defense Systems and later worked with telecom companies in Canada and the U.S. This experience included several years as an executive at Newbridge Networks, which was acquired by Alcatel in 1999. After "chasing the golden egg" with a technology startup for two years, Range signed on with net.com as vice president of operations.

Q: Net.com obviously is a survivor in an industry that has been hit really hard in recent years. What has it done right?
Range: Yes, the company is actually 21 years old this year. Its claim to fame is that, in the late 1980s and early 1990s, it was almost a de facto standard for providing wide-area networking equipment to link geographically remote organizations together. So if you were a global company that had facilities all over the world, our equipment was used to link those facilities together. These were the days when our customers would buy the networking equipment and lease lines from the telephone companies. That model shifted significantly in the mid '90s when the telephone companies realized they were missing out on a revenue opportunity and decided to buy the equipment and offer that service themselves to Fortune 1000 companies. Net.com's response was to shift its market focus away from enterprise space and toward the Tier One carrier space and the federal government. There is good news and bad news there. The bad news is that our development of Tier One-focused products began late in the '90s, so we did not experience the boom that many of our competitors did during the '90s; the good news is that, as a result of our focusing in on the federal government, we were able to sustain and even grow our business during the massive downturn that occurred from 2002 until today.

We are a $130m to $140m a year business and about 75 percent of that is with the federal government. We have three product platforms. The first, Promina, represents the lion's share of our business. We sell Promina to government agencies as well as to the military. A lot of the central command and communications systems that you see out in Afghanistan, Kuwait and Iraq are run over our equipment. The Hummers you see running around in the desert have our equipment on them. That has obviously kept us very busy.

We also have two other, newer products. One is called SCREAM, which stands for Service Creation and Manager. We position that with Tier One carriers for broadband aggregation so we can provide end-to-end quality of service in DSL broadband deployments. Our third product is SHOUTIP, which is a voice-over-IP product. Each of those products, from a supply-chain point of view, has very different challenges in terms of what we have to manage. Our operations strategy is all about being able to differentiate the way we manage our supply-chain depending upon the type of products, the type of demand and the type of supply that all have to get matched together.

Q: What was your biggest supply-chain concern when you joined net.com?
Range: Basically, what I found was a completely distributed approved vendor list, where we had a large number of vendors supplying a very small number of components to us. So the challenge was to get a lot of that stuff consolidated and to get a disproportionate amount of mind share from the key cost-driving suppliers and the key technology suppliers that we are using for our products. We had to do a lot of analysis to truly understand where we could get the biggest bang for our buck.

Then I had to look at the organizational model. We had a model that-for want of a better term-had abdicated responsibility for the supply chain to our contract manufacturers. From an organizational point of view, the only real interface we had to the supply chain at that time was directly with our contract manufacturer. So, for anything to the right side of our contract manufacturer, we had (1) no visibility; (2) no influence; and (3) no real understanding of what liability existed-nor of what the market trends were. So the first thing we did was to shift that model so as to identify direct relationships with the key technology and cost-driving vendors.

This was essential because we were trying to influence and drive costs out of our products. The way we did that was to leverage our design team. We said, 'Don't design in that vendor unless they are able to give us a break from a cost point of view.' So this is an ongoing model we have put in place that really leverages design and cost upfront, as opposed to assuming that you can drive cost reduction as a function of volume. We enter into annual negotiations with those strategic vendors every year and it is all around how do we leverage design and cost. We believe the time to negotiate cost is during the design phase, not the go-to-volume phase. So that is the first thing we did with our supply chain.

Q: And step two?
Range: The second thing was to try and deal with very spiky up-ticks in demand that we were experiencing 18 months ago and, at the same time, to cope with prioritizing our supply and demand picture based upon business rules. If you look at the maturity of the tools that are available today, specifically at what MRP does or does not do for you, you will see that MRP enables you to match supply and demand based purely on quantities, dates and lead times. But what we had to do was allocate supply to orders that had differing priorities. The reason that is an issue for us is that the federal government has a system of rated orders known as DX and DO orders. If you accept a DX- or DO-rated order from the federal government, those orders must take priority over any other commercial orders. So if we get a DX or DO order, we have to give it priority within our backlog.

The problem was we had no real way of managing those priorities other than playing around with the dates in MRP. The solution that came to the fore for us was Valdero's supply and demand match. Valdero enables us to look at our backlog at any given point in time and run a scenario based upon our business rules within any given period. So we are able to meet our contractual obligations with the federal government by prioritizing the DX and DO orders we receive above any other orders.

Q: Do you also use this solution to evaluate whether you should accept an order?
Range: We actually use Valdero to do order promising. Let's say you are in a situation where you have a backlog and you are balancing available supply with demand. At any given point in time a DX order or a higher priority order can come in and totally displace all of the supply- chain balancing that you have already done. That is exactly what we use the Valdero tool to do-to make sure that we understand the impact of a higher priority order on the overall supply chain.

Q: Was there a third initiative?
Range: Yes, we are now in the middle of our third initiative, which is looking at the information from the far left, the demand signal, to the far right, the vendor community. If you could map the information that is available to you from the start to the end of the supply and demand picture, it basically is a huge bell curve. From the net.com perspective, for example, we have a tremendous amount of information available about our backlog, a tremendous amount of information about the cost of our products, a tremendous amount of information about our inventory, etc. As I go further to the left, toward the demand side of things, or further to the right, toward the supply side of things, I have a lot less information. And with an outsourced model, when you have to work through a contract manufacturer, that information base gets significantly diminished. So, as part of our third initiative, we have two projects going. One is looking at how we can get a much more complete picture of our demand signal and the other is looking at how we can get a much more complete picture of our supply chain through the contract manufacturer.

Q: How do you go about that?
Range: What we do is slice and dice our demand signal into a three-by-three matrix. We stratify our part numbers into three different types of demand: Rogues, Runners and Repeaters. Clearly, Runners are items that we move every day, every week, every month. If we don't sell a lot this week, we will sell it next week. Repeater orders come in less frequently than Runners but nonetheless they do repeat themselves, perhaps once a month and definitely once a quarter. Rogues are demanded once in a blue moon and even more challenging is that they can be very spiky in nature. We might get an order this week for 100 units and then another order in six months for 10 units. Each of those three different types of demand needs a different supply-chain strategy.

Now, the other part of the three-by-three matrix addresses what we want the contract manufacturer to do with the demand signal. The way we have done that is to use the terms Fixed, Flex and Forecast. Anything that is Fixed, we want the contract manufacturer to go out and buy material and build it. We are prepared to take a 100 percent risk. If something is Flex, that means we want the contract manufacturer to position the material within the supply chain. If it is Forecast, we just want them to secure the long lead-time items.

So using the matrix of Rogues, Runners, Repeaters and Fixed, Flex and Forecast we are able to look at the amount of liability we would expect within the supply chain. The whole challenge for us is how we take that concept and translate it through the supply chain. So we are doing two things: (1) developing our own internal software to deal with the capturing and the management of this 3R/3F scenario, and (2) we are looking to some third-party vendors to see whether or not we can use their products to deal with this concept in the supply chain.

Q: Do you have any quantifiable results you can share?
Range: Before implementing the Valdero solution, we had a default lead time of 20 days and we only met that 30 percent of the time. We now have an on-time-delivery performance to our first promise date of around 85 percent. We have done away with the default lead-time because we basically drive to either a customer-requested date or a rated priority. We have reduced our inventory from around $19m to around $9m as a result of the visibility that we get, so we have been able to do away with big chunks of just-in-case inventory. We also are doing a much better job of prioritizing what we want the contract manufacturer to build in order to link up with our priorities.

Network Equipment Technologies Inc., doing business as net.com, develops and delivers platforms for broadband, internet protocol telephony and multiservice networks. An architect of the networking industry, net.com has been supplying service providers, governments and enterprises around the world with reliable networking technology for more than 20 years. Jeff Range, a native of the U.K., joined net.com in 2002. He began his career as an engineer with Marconi Space and Defense Systems and later worked with telecom companies in Canada and the U.S. This experience included several years as an executive at Newbridge Networks, which was acquired by Alcatel in 1999. After "chasing the golden egg" with a technology startup for two years, Range signed on with net.com as vice president of operations.

Q: Net.com obviously is a survivor in an industry that has been hit really hard in recent years. What has it done right?
Range: Yes, the company is actually 21 years old this year. Its claim to fame is that, in the late 1980s and early 1990s, it was almost a de facto standard for providing wide-area networking equipment to link geographically remote organizations together. So if you were a global company that had facilities all over the world, our equipment was used to link those facilities together. These were the days when our customers would buy the networking equipment and lease lines from the telephone companies. That model shifted significantly in the mid '90s when the telephone companies realized they were missing out on a revenue opportunity and decided to buy the equipment and offer that service themselves to Fortune 1000 companies. Net.com's response was to shift its market focus away from enterprise space and toward the Tier One carrier space and the federal government. There is good news and bad news there. The bad news is that our development of Tier One-focused products began late in the '90s, so we did not experience the boom that many of our competitors did during the '90s; the good news is that, as a result of our focusing in on the federal government, we were able to sustain and even grow our business during the massive downturn that occurred from 2002 until today.

We are a $130m to $140m a year business and about 75 percent of that is with the federal government. We have three product platforms. The first, Promina, represents the lion's share of our business. We sell Promina to government agencies as well as to the military. A lot of the central command and communications systems that you see out in Afghanistan, Kuwait and Iraq are run over our equipment. The Hummers you see running around in the desert have our equipment on them. That has obviously kept us very busy.

We also have two other, newer products. One is called SCREAM, which stands for Service Creation and Manager. We position that with Tier One carriers for broadband aggregation so we can provide end-to-end quality of service in DSL broadband deployments. Our third product is SHOUTIP, which is a voice-over-IP product. Each of those products, from a supply-chain point of view, has very different challenges in terms of what we have to manage. Our operations strategy is all about being able to differentiate the way we manage our supply-chain depending upon the type of products, the type of demand and the type of supply that all have to get matched together.

Q: What was your biggest supply-chain concern when you joined net.com?
Range: Basically, what I found was a completely distributed approved vendor list, where we had a large number of vendors supplying a very small number of components to us. So the challenge was to get a lot of that stuff consolidated and to get a disproportionate amount of mind share from the key cost-driving suppliers and the key technology suppliers that we are using for our products. We had to do a lot of analysis to truly understand where we could get the biggest bang for our buck.

Then I had to look at the organizational model. We had a model that-for want of a better term-had abdicated responsibility for the supply chain to our contract manufacturers. From an organizational point of view, the only real interface we had to the supply chain at that time was directly with our contract manufacturer. So, for anything to the right side of our contract manufacturer, we had (1) no visibility; (2) no influence; and (3) no real understanding of what liability existed-nor of what the market trends were. So the first thing we did was to shift that model so as to identify direct relationships with the key technology and cost-driving vendors.

This was essential because we were trying to influence and drive costs out of our products. The way we did that was to leverage our design team. We said, 'Don't design in that vendor unless they are able to give us a break from a cost point of view.' So this is an ongoing model we have put in place that really leverages design and cost upfront, as opposed to assuming that you can drive cost reduction as a function of volume. We enter into annual negotiations with those strategic vendors every year and it is all around how do we leverage design and cost. We believe the time to negotiate cost is during the design phase, not the go-to-volume phase. So that is the first thing we did with our supply chain.

Q: And step two?
Range: The second thing was to try and deal with very spiky up-ticks in demand that we were experiencing 18 months ago and, at the same time, to cope with prioritizing our supply and demand picture based upon business rules. If you look at the maturity of the tools that are available today, specifically at what MRP does or does not do for you, you will see that MRP enables you to match supply and demand based purely on quantities, dates and lead times. But what we had to do was allocate supply to orders that had differing priorities. The reason that is an issue for us is that the federal government has a system of rated orders known as DX and DO orders. If you accept a DX- or DO-rated order from the federal government, those orders must take priority over any other commercial orders. So if we get a DX or DO order, we have to give it priority within our backlog.

The problem was we had no real way of managing those priorities other than playing around with the dates in MRP. The solution that came to the fore for us was Valdero's supply and demand match. Valdero enables us to look at our backlog at any given point in time and run a scenario based upon our business rules within any given period. So we are able to meet our contractual obligations with the federal government by prioritizing the DX and DO orders we receive above any other orders.

Q: Do you also use this solution to evaluate whether you should accept an order?
Range: We actually use Valdero to do order promising. Let's say you are in a situation where you have a backlog and you are balancing available supply with demand. At any given point in time a DX order or a higher priority order can come in and totally displace all of the supply- chain balancing that you have already done. That is exactly what we use the Valdero tool to do-to make sure that we understand the impact of a higher priority order on the overall supply chain.

Q: Was there a third initiative?
Range: Yes, we are now in the middle of our third initiative, which is looking at the information from the far left, the demand signal, to the far right, the vendor community. If you could map the information that is available to you from the start to the end of the supply and demand picture, it basically is a huge bell curve. From the net.com perspective, for example, we have a tremendous amount of information available about our backlog, a tremendous amount of information about the cost of our products, a tremendous amount of information about our inventory, etc. As I go further to the left, toward the demand side of things, or further to the right, toward the supply side of things, I have a lot less information. And with an outsourced model, when you have to work through a contract manufacturer, that information base gets significantly diminished. So, as part of our third initiative, we have two projects going. One is looking at how we can get a much more complete picture of our demand signal and the other is looking at how we can get a much more complete picture of our supply chain through the contract manufacturer.

Q: How do you go about that?
Range: What we do is slice and dice our demand signal into a three-by-three matrix. We stratify our part numbers into three different types of demand: Rogues, Runners and Repeaters. Clearly, Runners are items that we move every day, every week, every month. If we don't sell a lot this week, we will sell it next week. Repeater orders come in less frequently than Runners but nonetheless they do repeat themselves, perhaps once a month and definitely once a quarter. Rogues are demanded once in a blue moon and even more challenging is that they can be very spiky in nature. We might get an order this week for 100 units and then another order in six months for 10 units. Each of those three different types of demand needs a different supply-chain strategy.

Now, the other part of the three-by-three matrix addresses what we want the contract manufacturer to do with the demand signal. The way we have done that is to use the terms Fixed, Flex and Forecast. Anything that is Fixed, we want the contract manufacturer to go out and buy material and build it. We are prepared to take a 100 percent risk. If something is Flex, that means we want the contract manufacturer to position the material within the supply chain. If it is Forecast, we just want them to secure the long lead-time items.

So using the matrix of Rogues, Runners, Repeaters and Fixed, Flex and Forecast we are able to look at the amount of liability we would expect within the supply chain. The whole challenge for us is how we take that concept and translate it through the supply chain. So we are doing two things: (1) developing our own internal software to deal with the capturing and the management of this 3R/3F scenario, and (2) we are looking to some third-party vendors to see whether or not we can use their products to deal with this concept in the supply chain.

Q: Do you have any quantifiable results you can share?
Range: Before implementing the Valdero solution, we had a default lead time of 20 days and we only met that 30 percent of the time. We now have an on-time-delivery performance to our first promise date of around 85 percent. We have done away with the default lead-time because we basically drive to either a customer-requested date or a rated priority. We have reduced our inventory from around $19m to around $9m as a result of the visibility that we get, so we have been able to do away with big chunks of just-in-case inventory. We also are doing a much better job of prioritizing what we want the contract manufacturer to build in order to link up with our priorities.