Category Archives: Forecasts

The Hierarchy of Supply Chain Metrics

Share This on Linked In

A classic article in the Supply Chain Management Review discusses “the hierarchy of supply chain metrics” because measurement is the cornerstone of operational success. According to the article, which builds on AMR Research, the best approach to measurement uses a three-tiered hierarchy that allows managers to quickly assess overall supply chain health at the top level, diagnose problems at the mid-tier, and identify corrective actions at the ground level.

At the top level, there are three core metrics that allow a manager to quickly gauge the overall health of the supply chain:

  • Demand Forecast Accuracy
  • Perfect Order
  • SCM Cost

At the next level, a composite cash-flow metric provides a diagnostic tool that allows a manager to zero in on the components that are likely the cause of any inefficiencies. The metric allows managers to determine whether there’s a balance between the time it takes suppliers and the time it takes customers to pay, whether the inventory metric (which can contribute to high costs) needs further analysis, and whether cash flow is being appropriately managed. The metric consists of:

  • Accounts Payable Turnaround Time
  • Inventory Totals
  • Accounts Receivable Turnaround Time

At the bottom level are the day-to-day metrics that measure performance across the different supply chain management activities and allow root-cause analyses when one of the higher-level metrics indicates a potential problem with efficiency or cost management. These metrics measure supplier effectiveness, operational effectiveness, and cost management effectiveness and include:

  • Supplier Quality
  • On-Time Delivery
  • Remaining Inventory
  • Purchasing Costs
  • Direct Material Costs
  • Production Schedule Variance
  • Plant Utilization
  • Work-in-Process Inventory
  • Order Cycle Time

Utilized properly, the hierarchy can lead to great success. However, companies can face significant challenges when designing and implementing measurement programs. The article offers seven recommendations for tackling the challenges that will arise:

  • Follow these four universal principles
    1. Keep it Balanced
    2. Work from the Outside In
    3. Focus on the Outcome
    4. Use the 80/20 rule and don’t choose too many metrics.
  • Proactively address organizational resistance
  • Beware of tunnel vision and ensure the metrics you choose address interactions and interdependencies
  • Analyze root causes when issues arise
  • Use a top-down approach to analysis
  • Measure enablers
  • Measure in the context of a performance-management program

Now That Your Demand Planning Strategy Is In Play, Improve It!

In our last post on the topic, we reviewed the story of how Linksys improved forecast accuracy at the SKU level by 350% with better demand planning, as told by Robert Bowman in Free The Enterprise! This emphasized the need to put a good demand plan in place and illustrated the importance of good demand planning strategies.

One key component of a demand planning strategy is a demand sensing strategy that will let you know when market conditions are varying from forecasted predictions in (near) real time, allowing you to update the forecast before you stock-out or, even worse, get stuck with thousands of units of obsolete inventory. The recent edition of the The VCF Report had a great article by Lora Cecere of AMR on Forecasting Recovery Strategies and “Seven Ways to Sense Demand and Predict the Upturn” that you should read to give you insight into how to tweak your demand sensing, and associated demand planning, strategies for best results.

Lora offered the following seven tips to help your company sense demand, and even predict the upturn, so that you can make timely decisions and reap the profits that will be yours for the taking, if you are ready.

  1. Make Better Use of Downstream Data from Retailers
    POS (Point-of-Sale) and inventory movement data can be used to shorten replenishment times.
  2. Implement VMI For Your Customers
    This will help you to better sense true demand and avoid stock-outs as you will have immediate access to channel wide data.
  3. Use Downstream Data from Distributors
    This will give you visibility into the reseller market and a better picture of overall demand.
  4. Move to Active Forecasting
    And update your forecasts weekly instead of monthly for short-life products and monthly instead of quarterly for longer-life products.
  5. Tap into Sales Contract Data
    This is critical for effective planning of make-to-order and configure-to-order supply chains.
  6. Actively Use Market Data
    Channel data and third-party data can be used to sense channel trends and predict when a certain product or service category is about to take off in the marketplace.
  7. Sense Service Requirements
    Link your demand-sensing activities with your strategic service management planning for better results.

For more details, see the article.

Now That You Have Your Demand Planning Strategy in Place, Use It!

In our last post, we reviewed Infor’s top ten demand planning strategies. In this post we’re going to illustrate why you need to put your demand planning strategy into action immediately using a recent case study from Global Logistics & Supply Chain Strategies as our example.

In Free The Enterprise! Bust the Silos in the Supply Chain, Robert Bowman tells us how, not too long ago, Linksys (the router division of Cisco Systems Inc.) had a record of only 20% accuracy and could only manage supply and demand for its top 200 SKUs. However, after taking appropriate actions, it was able to reduce inventory by 35%, backlog by 60%, and obsolete inventory by 40% in only twelve months. It also increased supplier fill rate from 65% to 95% while reducing expedited shipments more than tenfold from 40% down to 3%. And forecast accuracy at the SKU level increased 350% to a much more acceptable 70%.

How did Linksys do this? The Vice President of Operations tore down the silos between the demand forecasting and product management teams and created a formal S&OP organization that served as a data clearinghouse and a foundation for a company-wide demand forecast. No longer was forecasting a monthly spreadsheet exercise conducted in isolation by the demand forecasting team. In the new structure, a cross-fuctional forecasting team was formed that solicited input from finance, sales, marketing, purchasing, and supplier management before constructing a forecast. This is a much better situation than the one where no one trusted the forecast and sales would inflate its numbers to insure product availability.

In addition, the VP instituted an aggregated view of forecast, inventory, and production data for each SKU that he called “gameboards” supported by a an underlying software platform. No purchase was permitted unless it was based on factual information from the gameboard.

However, the effort ultimately succeeded because the internal walls were torn down and all organizational groups learned to work together in harmony and trust the forecast that was produced as a group. This is not always an easy effort. As the author astutely notes, “companies might find it easier to tear down functional walls separating them from external partners, such as suppliers, than those between internal departments“. In addition, “independent partners understand the need to work closely together, while individuals with a common employer tend to gravitate toward their immediate areas of responsibility“. And this tendency to stick to silos will continue unless you align incentives. If sales is incented to push product even if inventory is high and factories are incented to increase production even if the demand is not there, everyone ultimately loses. Everyone needs to be incented against the same goal and off of the same metrics, which must capture TCO reductions and ROI improvements. And the organization needs to move to a collaborative demand-driven mentality focussed on compliance with the agreed upon operational procedures.

Six Ways Companies Mismanage (Supply Chain) Risk

A recent Harvard Business Review article by Rene M. Stulz dives into “six ways companies mismanage risk” (membership required) that are just as applicable to supply chain operations as they are to financial operations. As the article points out, these missteps are just as likely to occur in good economic times as they are in the rough economic times we are currently experiencing, but rough times will magnify the impact of the mistakes considerably.

The six mistakes highlighted in the article are:

  • reyling on historical data
    Historical data is a starting point, not a destination. For example, look at how well real estate investment managers who assessed risk on the basics of statistics over the past three decades did in 2007. Closer to home, consider how well you would have done in your fuel hedges in early 2008 (before the price of oil dropped over 60%) or with your logistics hedges in late 2007 (before global shipping volumes were cut in half).
  • focussing on narrow measures
    Focussing only on-time deliveries misses the point. It’s about the perfect order — the right product of the right quality shipped using the right method with the right carrier at the right price delivered to the right customer at the right time. If you ship the wrong product, or the quality is insufficient, or you have to expedite it and it costs three times as much, you’re losing money and your metric will never capture the losses.
  • overlooking knowable risks
    Meticulous review and careful thought allows one to identify almost every possible risk, including risks in the instruments used to measure the risk. For example, if you are using an index to hedge against cost increases, and that index lags reality by three months, you could be cut off-guard by rapid cost increases or decreases due to unexpected supply or demand disruptions (caused by natural disasters, for example).
  • overlooking concealed risks
    Risk takers in your organization may deliberately hide risks that they feel are unlikely, and jeopardize an entire sourcing plan or production line. For example, if you’re in food, and your supply manager decides to source all of your tomato crop from coastal Florida because of volume-based cost savings, you’re at risk of an immediate supply disruption every time a hurricane sweeps up the cost.
  • failing to communicate
    If you can’t clearly explain the risks in your plan, systems, and organization, chances are they’ll be ignored, or at least severely underestimated. For example, if you’re assuming uninterrupted supply from a single-source supplier, and that risk goes overlooked, that could be a real problem in this economy.
  • not managing in real time
    Unless you’ve been hiding under a rock in a cave, you’ve probably noticed the volatility of the global markets lately, including supply volatility (as suppliers go out of business) and demand volatility (as customers reduce their spending).

All these mistakes will cost your dearly in the current economic climate, so its worthwhile reviewing your risk management strategy to make sure you haven’t made any of them. For more information on risk management, and best practices, see the risk management posts here on Sourcing Innovation.

Infor’s Top 10 Demand Planning Strategies

The bullwhip effect is as true today as it ever was in modern, elongated global supply chains where small errors at the front are magnified throughout the process.
Andrew Kinder, Director of Product Marketing, Infor

Forecasting is tough. Really tough. Especially in today’s market where consumers are fickle, credit is an unpredictable tide, and a single competitor innovation can completely change the market landscape. You have to forecast with foresight, balance judgmental and statistical methodologies, and focus on aggregate demands while continually sensing demand. And you have to be on your toes.

So how do you get it right? Although each situation has it’s own unique qualities, and any solution you acquire will have to have its model tweaked for your reality, there are some general steps that you can take that, if performed properly, will greatly increase your chances of success. These steps were captured quite nicely in a recent Infor top 10 checklist that was published last fall in an Industry Week article.

  1. Get the Process Right
    Demand planning is a sub-process within integrated business planning, not a stand-alone activity.
  2. Select the Right Level of Aggregation
    Do you aggregate demand by product family or geographic region? Why?
  3. Collaborate
    Statistics provides a foundation to build on, but the real value comes from over-laying expert knowledge that a system cannot know and cannot infer, such as a new marketing effort or an announcement by your competitor that was taken negatively by the market.
  4. Influence Demand
    Use coordinated marketing events and promotions to swing the forecast into favorable territory.
  5. Measure
    Select the right set of linked key performance indicators and measure against them regularly. This will tip you off to demand swings and allow you to tweak the forecast before it becomes a problem.
  6. Educate
    Before allowing someone to provide input into the forecast, it is critical that they understand how their contribution will impact the forecast and the performance against the demand plan. Otherwise, they may just guess and provide bad input that instantly ruins your best efforts.
  7. Cleanse
    Good, clean, data is an absolute.
  8. Manage By Exception
    Remember that 80% of your return can be achieved by actively managing only 20% of the forecast.
  9. The Error Term is Your Safety Stock
    A good statistical forecast will have an appropriate error which drives an appropriate safety stock target.
  10. Deploy a Proven Best-of-Breed Technology Solution
    According to Aberdeen, companies that excel in demand management are two-and-a-half times as likely to have implemented a best-in-class demand planning system.

All-in-all, it’s a great demand-planning checklist.