We're excited to present a new online feature: the Supply Chain Leader Virtual Roundtable. Created to feature the latest thinking from the world's preeminent supply chain thought leaders, the Supply Chain Leader Virtual Roundtable will focus on a different area of supply chain management each month. Inventory optimization serves as the first topic of discussion, and our contributors include:
 |
 |
 |
Ramesh Raj, Enterprise Architect, Global Supply Chain, Dell |
Noha Tohamy, Vice President, Head of Research, AMR Research |
Adeel Najmi, Vice President, Product Strategy and Planning, i2 |
SCL: We have heard a lot of buzz about inventory optimization in the last five to seven years. Can you separate the hype from the reality? How much have companies really benefited from this technology?
Ramesh Raj, Dell: Inventory optimization rightfully captured a lot of attention from executives in the last several years. There is some level of hype about the sophistication of the math, etc. In my opinion, however, the value is not just in the math itself, but rather in how the insights from inventory optimization can be applied to focus people from different functions on the right trade-offs and then scale across the entire organization. Therefore, this technology must be easy to use and scale globally.
Noha Tohamy, AMR: Well, multi-echelon inventory optimization is truly a powerful tool. It allows companies to determine the right inventory targets based on their supply and demand across the network. This typically results in lower inventory targets and/or improvement in service levels. It is not unheard of to reduce inventory levels by 10 to 25 percent while improving or maintaining customer service levels. Also, with better inventory decisions, a company can typically save on expedited transportation costs, which is usually a benefit.
Raj: Companies can benefit from inventory optimization in many ways. This technology can help align inventory mix to enable better service levels, whether it is service levels of parts supply to factories or end-products to customers such as Wal-Mart. The technology can indeed reduce inventory overall, which translates to dollar savings. In our experience, we have seen potential for benefits with inventory reduction anywhere from 10 percent to more than 50 percent while at the same time significantly improving in-stock levels and reducing excess inventories—this is the beauty of better mix alignment.
Tohamy: Despite clear benefits such as this, the adoption of inventory optimization is still very limited. How can such a powerful tool not be more widely adopted? I think this has to do with a few factors including the lack of executive sponsorship and the lack of alignment between the goals of the tools and the organizational structure. So even though on a case-by-case basis the benefits are astounding, companies have yet to widely adopt inventory optimization and therefore benefit from this technology.
Adeel Najmi, i2: i2 is fortunate to have worked with supply chain leaders that were already known for operational excellence. Customers have often reported quite amazing results from inventory optimization at i2 Planet and Directions conferences. Let me mention just three of these here: Southwest Airlines reported a $5 million benefit within the first six months of implementing i2 Inventory Optimization in its service parts business. The airline projected a ten-year expected cash flow benefit exceeding $30 million. Texas Instruments reported a 25 percent improvement in on-time delivery in the business line where it had implemented i2 Inventory Optimization. Given this initial success, Texas Instruments has rolled out this technology to the rest of the company. PepsiCo reported using i2 Inventory Optimization to fine tune inventory settings and balance strategy, tactics and cash utilization.
Technology plays a critical but small part of these and numerous other successes we have seen with our customers, however. People and process have the more important roles.
At i2 we have a very formal methodology of identifying and quantifying benefits we believe our customers will achieve. But clearly, if the requisite process and mindset changes do not happen—and people continue the same old practices even with new systems—actual results may be lower than we estimated. So, while there is plenty of truth in the hype, the benefits from simply better math calculations are minimal. The real value lies in companies putting these technologies to work in their businesses every day.
SCL: How has the practice of inventory management evolved over the last few years?
Raj: I would say that inventory management practices at Dell have evolved over two stages. Initially we focused on providing visibility into our needs and demand forecasts to our suppliers while maintaining very simple, measurable and easy-to-execute, across-the-board inventory norms and policies. The rest was left to the suppliers. Over time, we built on this approach by increasing the collaboration processes and tools between Dell and our suppliers. In a pure configure-to-order direct business model, this worked well.
More recently, our business model has changed significantly. For example, we have entered the retail channel. We now need to build some inventories based on forecasts just in time to meet the retail and distribution demands. With these changes, we have also needed to change our inventory management processes and enabling technologies. We now have the ability to work with differentiated inventory policies that account for supply and demand risk across our portfolio.
Tohamy: When we talk about inventory optimization now, we are talking about multi-echelon inventory optimization. This is different from the traditional single-echelon, sequential approach in which the demand is forecasted for each of the echelons separately or forecasted for the distribution/customer-facing nodes and then passed up to the echelon above.
Multi-echelon inventory optimization solves one problem, taking into account the demand forecast of the lowest echelon. Demand forecasting and inventory replenishment decisions are made at the enterprise level in a single optimization exercise rather than in a sequence of sub-exercises for each echelon. Multi-echelon inventory optimization takes into account lead time and lead-time variations in all the higher echelons. The big difference of this approach is that companies don't estimate what they should already know. And because of that, they minimize the forecast error and the bullwhip effect.
Another advancement is demand-driven inventory optimization. This means that I am eliminating the latency of using my customer demand as the signal for how much inventory to position. Instead of relying on historical forecasts, I can leverage actual sell through data to decide how much inventory each node needs.
Najmi: In the last few years, inventory management practices have been radically changed at a handful of visionary companies that have deployed breakthrough innovations in both process and technology. Here are some of the most important changes I have seen:
- Most companies have moved away from a simple "days of cover" based inventory approach and toward developing inventory policies for "right sizing" inventory for supply and demand risk.
- As Noha explained, many companies are moving from single-echelon to multi-echelon inventory optimization.
- Until recently, leading companies would use simple ABC segmentation, typically by sales volume, to manage target service levels across their portfolio. Now, the savviest companies are effectively developing "designer supply chains" by tailoring inventory strategies for each distinct portfolio segment. They are using sophisticated micro-segmentation technologies to dissect their portfolio along many dimensions such as volume, revenue, margin lifecycle stage and so on.
- Until about five years ago, most companies reviewed inventory policies one to four times a year. Supply chains have become very dynamic and complex, however. Working with pioneering customers such as Texas Instruments, we have actually seen tremendous value from inventory policy checks as frequent as once a week in many industries. When the solution suggests a significantly different target compared to previous weeks for a specific item stock, the first question inventory planners ask is, "What has changed and why?" The insights our customers gain from root-cause analytics enabled by our tools are even more valuable than the mere automation capability to keep inventory mix and positions in line with the dynamic supply chain.
- We are also seeing a growing trend toward the use of deep operational analytics where companies conduct regular cross-functional reviews of inventory performance. Our inventory optimization tool enables root-cause analysis of why there were stock-outs or excess inventory in a given situation. For example, in one case last year, the insights gained from this kind of postmortem enabled a customer to achieve results that were more than five times better than our original estimates.
As Noha mentioned previously, the adoption of inventory optimization is still very limited. My estimate is that less than five percent of the companies that can benefit from these best practices have actually achieved this level of process maturity. This is why, as solution providers, we continue to build detailed "process playbooks" to help companies achieve organizational alignment and facilitate change management, which is complemented by an investment in simple, easy-to-use enabling workflows so that more companies can realize greater efficiencies in their supply chains.
SCL: With the economic downturn, there is a resurgence of interest in inventory management. What role can inventory management have for companies looking for strategies to cope with the downturn?
Raj: Better inventory management can help reduce cost by reducing cost of goods sold and operational expenses. This simply offers one additional way to improve our overall bottom line.
Najmi: When the water level in a river is lowered, all of the rocks become exposed. This is what is happening with inventory in the downturn. Companies that were meeting supply chain needs at the expense of excess inventory and sluggish inventory management practices ended up with a glut of inventories. i2 did an analysis from recent 10Q reports to look at inventory and demand trends during the last several months. Our analysis showed that companies such as Texas Instruments were able to draw inventory down quickly when demand tanked. As a result, Texas Instruments tended to maintain relatively steady inventory levels as a percent of sales. During the same period, other companies in the same industry showed a sharp rise in inventory levels as a percent of sales. Clearly, a company's level of excellence in inventory management can make direct impact on both the top and bottom line in today's environment.
Tohamy: Controlling working capital is important regardless of economic conditions. But, yes this recession has made it more imperative for companies to have better visibility into their inventory and try to do more with less. Companies wanted to do that intelligently, not slash inventories and risk jeopardizing customer service levels. Inventory optimization tools can certainly help with that. Many of the rules of thumb that companies have employed in the past for figuring out their inventory levels are no longer valid given the drastically changing demand picture. Also, with many suppliers facing financial risks, an inventory optimization tool can help a company determine the right inventory levels given an increasing supply variability and risk.
SCL: What are some of the short- and long-term actions companies can take to improve inventory performance?
Raj: In the short term, one can take a careful look at inventory policies. Clearly, since the business climate has changed, our policies must also be re-aligned to the new reality. We need to look at how we are positioning inventories to cope with today's uncertainty and fluctuations in demand. In the long term, we need to focus on putting the right processes and technologies in place for a sustainable advantage in overall performance. We should be asking questions such as, "What if the situation was to flip and we have resurgence in demand?" and "Will we be ready with processes and enabling tools to harness that opportunity when it comes?"
Najmi: I agree with Ramesh, the most effective approach I have found that can bring significant results is to make a formal review of inventory policies and evaluate alternative segmentation approaches and their impact on service levels and requisite inventory investments. With only a few weeks of analysis, companies can re-align their supply chain strategy to the new reality.
Another analysis that I highly recommend is to look carefully at forecast errors as well as bias. If forecasts are running consistently higher or lower than actual then it is an indication of systemic process breakdowns that could be fixed relatively quickly. A small improvement in the forecasting and demand consensus process may drive huge efficiencies in terms of inventory.
Finally, I would also echo Ramesh's comments about looking at this downturn as an opportunity to make fundamental process and technology improvements so that you are ready with a competitive advantage when demand picks back up again.
Tohamy: Usually the first step is to do an audit to determine the current levels of inventory. Second, companies should try to figure out why these levels are where they are. Is that because of a customer requirement? Is it because this is how they did it historically? Next, they should conduct analysis to see how much inventory is needed to meet their customer requirements and how they can save using an optimization tool. But before assuming that these savings are achievable, they have to take a critical look toward their organization and see if such a tool will be leveraged by their users. They have to make sure that users' incentives are aligned with the company goal of optimizing inventory and customer service levels across the board, not just for one region or one business unit.
Companies also need to take inventory optimization tools for a test drive. The strength of the model is very important, but the usability and transparency of the tool is equally important. The users have to be convinced that the model is doing the right thing.
SCL: What performance metrics are the best and most succinct to track progress in inventory management?
Raj: At Dell, whenever we talk about metrics, we ask how we can make the metrics actionable. So, we look at how we will track excess and short inventory situations and enable easy communication across all stakeholders—including our suppliers. These metrics must be easy to communicate, easy to track and easy to act upon. In parallel, companies must have an understandable and transparent way of compiling performance scorecards so that process owners can collaboratively learn to improve performance over time. At Dell, we use a simple color-coded scheme for tracking and monitoring inventory levels. This scheme is reflected in our tools and dashboards.
Tohamy: AMR uses a hierarchy of metrics to keep track of the general health of the supply chain. The idea behind this is that companies have to look at cost and service to understand how well or how poorly they are managing their supply chain. Companies need to look at inventory turns and inventory levels in conjunction with other key performance indicators such as customer service levels, transportation costs and forecast accuracy.
Najmi: Metrics, when correctly used, are critical for at least two reasons. First they drive people's behavior. Second, they provide an organization with the visibility necessary for the "check" step in the plan-do-check-act cycle. Good metrics are the foundation of good governance. Different overall objectives require different metrics; one size does not fit all.
While we cannot prescribe a single set of metrics that will work in all situations, here are some simple guidelines that can help companies build a metrics framework that works for their business situation:
- Identify no more than two or three top-level metrics that are aligned with overall corporate priorities. Too many metrics confuse the organization and dilute their effectiveness. On the other hand, having undue emphasis on a single metric usually has undesirable side effects when related trade-offs are neglected. For example, one cannot look at service level without also looking at the cost expended to sustain it.
- Do not neglect to understand the metrics your customers will use to measure you. Internally focused metrics alone will not be sufficient.
- Use metrics in combination with micro-segmentation to derive insights about which segments of your portfolio are underperforming and why.
- Include metrics that close the loop between planning and execution. If actual performance deviated from plans, ask, "When did you first know?" "When could you have first known?" and "What can we do to detect these same situations more proactively next time?"
- Keep it simple. Every metric should be clear and understood by the entire organization. An ambiguous metric will most definitely result in people stretching the definition to sugarcoat their report card as needed.
Clearly, in today's economic environment, everyone's attention has turned back to inventory. I think this discussion has shown us that there is no silver bullet, but there is a long-term, sustainable advantage to be achieved by taking a hard look at the processes and technology around inventory optimization. I thank Ramesh and Noha for their valuable contributions to this discussion.
Interviews were conducted by Lauren Bossers, Supply Chain Leader editor.
We want to hear about the issues and opportunities you're facing with regards to inventory optimization. Please share your thoughts in the comments section.
