Even the best laid plans can go off track if something changes unexpectedly- a major customer cuts orders, unexpected defects occur, requiring ad-hoc reaction, rather than careful, deliberate, advanced planning.

There are two major directions to approach inventory management from-- Top-Down and Bottom-Up. Most successful companies use a combination of both.

• Top-Down — this is the “macro” approach. Start with a goal, objectives, ABC (Pareto) analysis of estimated or historical usage, knowledge of overall processes and lead times. Set overall targets, by business unit at a minimum, preferably at a lower level, so that middle managers or even individual supervisors, work teams or administrative control personnel might be held more accountable. It takes more effort as the control is moved to a lower level.

Establish a tracking system, such as actual inventory versus target level. Compare numbers to actual sales, forecast. Monitor commitments and production plans against targets... Hold managers accountable for results and make them come back with reasons why targets cannot be met and solutions to the problems. Motivate them to solve underlying problems. Help them with problems outside of their scope of authority.

Another good tracking tool is Input-Output Control. Simply build a time-phased table of planned starting and ending inventories, showing starting, input, output and results. Then task employees to make the “delta’s” happen and track the actual values per period.

• Bottom-Up—Look at each item- determine cost, lead times, supply and demand reliability/variability, defect rate, transportation, storage, set-up/batch size considerations, buffers, process, handling considerations. Then set the proper planning methods and control parameters, to either default down from the enterprise, product line, commodity or department level to default down, or just establish them at the item/part level.

This takes a lot more effort than merely exercising Top-Down control, but it can deliver better results.

Educate and train people in inventory management and control approaches.

How to Control Inventory

After you do all your research and analysis, set targets and establish your control system, then you get to the hard part - actually making it happen.

Quick hits - Simply establishing the aggregate targets, understanding drivers, educating and training, setting up responsibility, establishing accountability and tracking results usually has significant effects. I have seen greater than 50% reductions from this alone. This can be the cheapest, fastest way of making some change happen, but it has a limited effect, because the approach lacks detail and won’t make major permanent changes in the ways that the business works without additional actions.

What is “Control?” - Control means to make something happen or to know why if it doesn’t, so that something might be done about it. Using that definition, there is no such thing as an uncontrollable situation. Someone once told me that he couldn’t control service inventory, because of unreliable vendor lead times. Nonsense! Unreliable lead times might be controlled by several strategies, such as: multiple sourcing, re-sourcing, safety stock, exhorting supplier to improve performance, ordering sooner, improving your own planning and reaction times, changing designs, alternate routing, training customers to order differently, having vendors stock raw materials. At least some of these would work in almost any situation.

Pitfalls of using control parameters

Wi

This paper covers how to set up and maintain Aggregate Inventory Management for improved investment and operations management. It is a “macro,” top-down approach that complements a company’s “micro” SKU (part number) level management techniques.

Definition, Goal and Objective

• Definition -- the APICS Dictionary defines Aggregate Inventory Management as ”Establishing the overall levels of inventory desired and implementing controls to ensure that individual replenishment decisions achieve this goal.”

It includes:


• How to assess overall investment levels and set targets.
• How to identify inventory investment level “drivers” and help control them
• How to link aggregate inventory management “macro” strategy to “micro” controls and develop accountability
• Performance measurements
• Specific techniques, such as ABC analysis, control parameters, inventory buildup charts, and input-output control.

• Goal -- Helps manage assets and make money.

• Objective -- Optimize inventory levels within the parameters of service, cost, logistics, process and investment objectives/constraints. Inventory management should be exercised to keep the lowest level of inventory consistent with achieving the objectives. Too much inventory reduces Return on Investment and Return on Assets (lower profits). It also tends to increase expenses, in the form of interest payments, handling and storage, management, damage, loss, obsolescence, tracking, taxes, insurance, etc.

Although most managers, accountants and taxing authorities regard inventory as an asset, treating it as such for operational purposes may create liabilities. You have probably heard stories about factories working to “keep people busy” or maximize “efficiency” and other similar nonsense. If they are making inventory that is not needed now, they are often wasting money. If they work just to keep people busy, they are still consuming material, energy and other resources that may not earn adequate profits. They may use resources that could better be used for more immediate and profitable needs. If inventory is deployed improperly, it may create liabilities. A customer of one of our clients had branch managers who would “hoard” products at their remote branches so that they “wouldn’t run out.” This created an excess of material in the wrong places.

How to Assess Inventory Investment Requirements

Survey

First, understand market, customer needs and service expectations; your own company needs, expectations, process, abilities; supplier abilities and mindset; industry norms and mindset; world-class best practices.

From this, you should learn how fast and reliably customers expect to get their shipments, what is involved to get raw materials and production completed, what the best in the industry are doing and plan to do, and what might be possible. For instance, if all competitors are shipping from stock, then you will either need to duplicate that feat, or determine how to manufacture very fast, or convince customers that your product is so great or so cheap that it is in their interest to wait while you make it to order. Or, you might figure out how to procure better or manufacture better in a way that allows you to carry less inventory.

The result of this step is to establish what industry inventory standards might be and what is possible. Make sure you have an “apples-to-apples” comparison: there may be significant differences among companies.

Measure Current and Historical Inventory Levels and Performance

Measure current and historical company inventory levels and performance, not just overall statistics, but broken down into levels of responsibility, commodity, area, type (raw material, work-in-process, finished goods, consignment) and market. Do this to help isolate figures down to levels of accountability and to show inventory investment performance by market, process or even product line. You may find that your systems are unable to do that, meaning that it is past time to make changes to them, whether that be to replace them, modify them or put in separate inventory tracking and control systems (recommended as a last resort).

The result of this step is to establish how your own company is doing and has been doing with inventory management.

Establish Performance Metrics

Establish performance metrics - Inventory is usually measured in currency value, such as U.S. Dollars ($USD). Another, complementary way is to measure it in velocity. For example, you might measure it in “turns” which relates to how many times it moves or “turns over” per year. For example, if there was an average of $100 in inventory in the last year and annual cost of sales for the last year was $2000, that would be calculated as cost of sales ($2000)/average inventory ($100)= 20 turns.

More turns (or “turnover”) is usually good, provided that cost, service or quality aren’t unacceptably affected. If they are, the answer is not simply to increase inventory, but to try to improve the underlying “drivers” influencing it instead, if possible and cost-effective. There are variations of the turnover (this term should not be confused with the European “turnover,” which usually refers to total sales for a period) formula, mainly in addressing how to calculate average cost of goods sold or inventory.

Sometimes, turns are calculated by comparing full sales value with average inventory cost or even equivalent sales value. To maintain easily comparable figures, state all numbers in fully “burdened” costs, using industry standard overhead/burden calculations, unless this is contrary to the standards of your industry or locality.

It is becoming more common to measure inventory performance in days coverage instead of turnover. People seem to relate to it better.

Inventory and sales may also be commonly measured in more industry-friendly terms, such as tons (steel), bushels (corn), housing units (construction or real estate) or ounces (gold).

A further refinement is to stratify the inventory by “Quality,” as asserted by Gary Gossard of IQR International. The idea of classifying inventory as active, slow-moving or obsolete has been around for a long time. Constantly track it, to highlight any change in inventory quality or condition, such as a new requisition for an item which is already in excess or obsolete. The active, weighted “good” inventory not exceeding your “days coverage” target, divided by the total inventory, multiplied by 100, it equals the Inventory Quality Ratio (IQR) number. 33-40% is typical for mediocre companies. 66% is considered pretty good.

All of these numbers can be time-phased, to show changes over time, due, for example, to seasonal supply and demand changes, or planned improvements. These can then be applied in still more detail to the appropriate organizations, product lines, trade channels, warehouses, planning groups or other responsible entities and then monitored for results.

The numbers should be capable of being “drilled” down or up, from the entire enterprise level to an individual SKU (Stock-Keeping Unit) transaction or part number. Managers or employees should be able to look at total figures for their areas of responsibility and readily identify specific problem areas down to lower levels and finally to specific items, policies, orders and decisions that accounted for them.

Here are typical Inventory System Metrics, which should be broken down by organization/responsibility, area, type, commodity, market/product, and time phased, with targets and actual values:

• Inventory Turnover or Days Coverage
• Inventory value or other unit of measure, such as tons
• Inventory “Quality,” including IQR and summaries of amounts of each type
• Customer service level, expressed how the CUSTOMER perceives it

ABC Analysis

Perform an ABC analysis, a simple, common and powerful tool for inventory management. It is based on Pareto’s law of “80-20.” The most common approach is to calculate demand in units, preferably for future periods, then calculate the total usage value at cost for each item (total cost of sales multiplied by units required) for a given future period. If future demand data are not available, the next best thing is to use history, but this won’t work well for items with major swings in demand over time. Sequence these in descending value. Typically, the top 10 to 15% of items account for 75-85% of value (“A” items), the next 20-30% account for 10-20% of value (“B” items) and everything else accounts for the rest, about 60-70% of the items, usually about 5% of the total value (“C” items). Your inventory should be less than these percentages for the “A” items, because they are much more tightly controlled and a little higher for B’s and significantly higher for C’s.

Then compare the list to actual values in inventory, plus actual and planned commitments. The answers will often suggest immediate corrective actions!

An ABC list suggests what to concentrate on to control most of the inventory investment. What it doesn’t tell you is that being short of a $.10 screw might prevent the shipment of a $5,000,000 radar unit, so ensure that there are control systems for all items, just control the expensive ones much more carefully. Err on the side of caution for the cheaper items, allowing a safety stock coverage or “two bin” approach to avoid stock outs, but keep inventory from getting out of control.

Create an Inventory Buildup Chart

Another good analysis tool is the inventory buildup chart. Use a standard x-y coordinate chart. Plot the cost build-up over time, by product group, with cost on the “y” (vertical axis) and time on the “x” (horizontal) axis. Normally, raw material cost accumulates first over time, followed by labor and overhead application. Allow for safety stocks, lot size inventory, transit stock, defects/rework/scrap, and normal finished goods and distribution pipeline stocking. Show the affect of consignment arrangements. Some people also treat accounts receivable as sort of a de facto inventory, until it is paid for. Once this chart is completed, show it around for shock value. Presented correctly, it will really make people think about the effect of constraints and decisions (just another form of constraint) on inventory. Then, work on changing the rules!

One company had a 14 month buildup curve, which was reduced to 4 months. At another company, the longest lead time material item accounted for only 20% of the product cost, so stocking only that item, instead of finished goods or instead of only reacting to orders, enabled them to radically reduce the response time for orders by 70%. It also added the flexibility of being able to use that raw material to make a number of different end items.

How to Identify and Control Inventory Drivers

Inventory drivers are things that tend to make inventory go up or down. Identify them and you will have some clue of why inventory changes. Understanding them is the beginning of gaining control. I’ve stated things that would drive inventory up, e.g.: more SKU’s. I refrain from stating the obvious: doing the opposite would reduce inventory. e.g.: reduce SKU’s to reduce inventory.

Key Drivers are covered briefly, as follows:

Number of SKUs

The more items you have, the more inventory you will need, in most cases. If you sell 500 widgets a year of A, then replace it with 250/year of A and 250 of B, you will probably need to carry more inventory. Why: demand and supply variability and total economic order quantities are likelier to be higher for 2 items than for one.

The more SKU’s in a product, the harder it is to bring matched sets of parts together at the same time. Because there are multiple items, with multiple vendors, kept and routed through multiple places or paths, with more opportunity for delays, defects, etc, more inventory will be needed.

The more operations there are and the longer that they take, the more inventory you will tend to have. More operations mean a longer supply chain. It may also mean differing lot sizes per operation and more places for delays and defects to occur. Process simplification helps reduce inventory.

The more facilities that inventory passes in and out of, the further apart those are and the harder they are to reach and pass material in and out of, the more inventory you will tend to have.

The more times inventory passes from the control of one system or organization to another and the less efficient the transfer is, the more inventory you will tend to have.

Lot/Batch Sizes

Lot/batch sizes greater than customer order delivery sizes tend to increase inventory. If customers order a product one at a time, but economics, handling or process considerations suggest that you make 1000 at a time, then you will have more inventory available than will be consumed per order, resulting in an accumulation of inventory. If you need to order things in cases, dozens, carloads, tons or weeks’ supply, but they are needed downstream in the supply chain in smaller increments, you will tend to accumulate more inventory.

The longer the lead time, the more inventory you tend to have. If something takes 16 weeks to get instead of 16 days, there is more inventory needed in process to cover the “pipeline” time. Whether it belongs to you or your vendor, it is increasing somebody’s cost, which ultimately will affect your cost and your customer’s cost. Longer lead time also means more chance of running out or having something go wrong out while waiting for it, which is usually dealt with by having additional inventory.

Carrying cost

This refers to the cost of owning inventory. For a closer look at these costs and the topic of inventory reduction, please see my article "Inventory Reduction - A How to Guide".

How to set Inventory Targets

After considering the current situation, drivers, and external situation, estimate what inventory levels should be, given certain sets of circumstances. There are impressive supply chain modeling tools to help you do this. Our experience is that developing an accurate detailed inventory behavior model is quite a chore to create and a major task to maintain, so we usually don’t. Normally working on projects with limited budgets, we study past behavior and focus on the main drivers, seeking to change a few with the greatest potential impact to achieve assigned objectives- sort of a “delta’ approach.

Don’t let us talk you out of sophisticated modeling tools, though. They have their place. When there are very large amounts of money involved and/or tricky constraints to work around, modeling tools will sometimes help. Many of the detailed control methods presented below contain elements of modeling.

Warning: Calculating or modeling inventory behavior solely by using the rules and parameters will nearly always be wrong. Why: If, for example, you assume that inventory will be an average of ? times the order quantity plus safety stock, you’ll most often be wrong. Actual supply and demand variability will differ. Defective items/customer returns may result in buildup. Unmatched sets of parts due to shortages will result in buildup. Generally, it is higher than the model would indicate.

Even the best laid plans can go off track if something changes unexpectedly- a major customer cuts orders, unexpected defects occur, requiring ad-hoc reaction, rather than careful, deliberate, advanced planning.

There are two major directions to approach inventory management from-- Top-Down and Bottom-Up. Most successful companies use a combination of both.

• Top-Down — this is the “macro” approach. Start with a goal, objectives, ABC (Pareto) analysis of estimated or historical usage, knowledge of overall processes and lead times. Set overall targets, by business unit at a minimum, preferably at a lower level, so that middle managers or even individual supervisors, work teams or administrative control personnel might be held more accountable. It takes more effort as the control is moved to a lower level.

Establish a tracking system, such as actual inventory versus target level. Compare numbers to actual sales, forecast. Monitor commitments and production plans against targets... Hold managers accountable for results and make them come back with reasons why targets cannot be met and solutions to the problems. Motivate them to solve underlying problems. Help them with problems outside of their scope of authority.

Another good tracking tool is Input-Output Control. Simply build a time-phased table of planned starting and ending inventories, showing starting, input, output and results. Then task employees to make the “delta’s” happen and track the actual values per period.

• Bottom-Up—Look at each item- determine cost, lead times, supply and demand reliability/variability, defect rate, transportation, storage, set-up/batch size considerations, buffers, process, handling considerations. Then set the proper planning methods and control parameters, to either default down from the enterprise, product line, commodity or department level to default down, or just establish them at the item/part level.

This takes a lot more effort than merely exercising Top-Down control, but it can deliver better results.

Educate and train people in inventory management and control approaches.

How to Control Inventory

After you do all your research and analysis, set targets and establish your control system, then you get to the hard part - actually making it happen.

Quick hits - Simply establishing the aggregate targets, understanding drivers, educating and training, setting up responsibility, establishing accountability and tracking results usually has significant effects. I have seen greater than 50% reductions from this alone. This can be the cheapest, fastest way of making some change happen, but it has a limited effect, because the approach lacks detail and won’t make major permanent changes in the ways that the business works without additional actions.

What is “Control?” - Control means to make something happen or to know why if it doesn’t, so that something might be done about it. Using that definition, there is no such thing as an uncontrollable situation. Someone once told me that he couldn’t control service inventory, because of unreliable vendor lead times. Nonsense! Unreliable lead times might be controlled by several strategies, such as: multiple sourcing, re-sourcing, safety stock, exhorting supplier to improve performance, ordering sooner, improving your own planning and reaction times, changing designs, alternate routing, training customers to order differently, having vendors stock raw materials. At least some of these would work in almost any situation.

Pitfalls of using control parameters

Wit

Measure Current and Historical Inventory Levels and Performance

Measure current and historical company inventory levels and performance, not just overall statistics, but broken down into levels of responsibility, commodity, area, type (raw material, work-in-process, finished goods, consignment) and market. Do this to help isolate figures down to levels of accountability and to show inventory investment performance by market, process or even product line. You may find that your systems are unable to do that, meaning that it is past time to make changes to them, whether that be to replace them, modify them or put in separate inventory tracking and control systems (recommended as a last resort).

The result of this step is to establish how your own company is doing and has been doing with inventory management.

Establish Performance Metrics

Establish performance metrics - Inventory is usually measured in currency value, such as U.S. Dollars ($USD). Another, complementary way is to measure it in velocity. For example, you might measure it in “turns” which relates to how many times it moves or “turns over” per year. For example, if there was an average of $100 in inventory in the last year and annual cost of sales for the last year was $2000, that would be calculated as cost of sales ($2000)/average inventory ($100)= 20 turns.

More turns (or “turnover”) is usually good, provided that cost, service or quality aren’t unacceptably affected. If they are, the answer is not simply to increase inventory, but to try to improve the underlying “drivers” influencing it instead, if possible and cost-effective. There are variations of the turnover (this term should not be confused with the European “turnover,” which usually refers to total sales for a period) formula, mainly in addressing how to calculate average cost of goods sold or inventory.

Sometimes, turns are calculated by comparing full sales value with average inventory cost or even equivalent sales value. To maintain easily comparable figures, state all numbers in fully “burdened” costs, using industry standard overhead/burden calculations, unless this is contrary to the standards of your industry or locality.

It is becoming more common to measure inventory performance in days coverage instead of turnover. People seem to relate to it better.

Inventory and sales may also be commonly measured in more industry-friendly terms, such as tons (steel), bushels (corn), housing units (construction or real estate) or ounces (gold).

A further refinement is to stratify the inventory by “Quality,” as asserted by Gary Gossard of IQR International. The idea of classifying inventory as active, slow-moving or obsolete has been around for a long time. Constantly track it, to highlight any change in inventory quality or condition, such as a new requisition for an item which is already in excess or obsolete. The active, weighted “good” inventory not exceeding your “days coverage” target, divided by the total inventory, multiplied by 100, it equals the Inventory Quality Ratio (IQR) number. 33-40% is typical for mediocre companies. 66% is considered pretty good.

All of these numbers can be time-phased, to show changes over time, due, for example, to seasonal supply and demand changes, or planned improvements. These can then be applied in still more detail to the appropriate organizations, product lines, trade channels, warehouses, planning groups or other responsible entities and then monitored for results.

The numbers should be capable of being “drilled” down or up, from the entire enterprise level to an individual SKU (Stock-Keeping Unit) transaction or part number. Managers or employees should be able to look at total figures for their areas of responsibility and readily identify specific problem areas down to lower levels and finally to specific items, policies, orders and decisions that accounted for them.

Here are typical Inventory System Metrics, which should be broken down by organization/responsibility, area, type, commodity, market/product, and time phased, with targets and actual values:

• Inventory Turnover or Days Coverage
• Inventory value or other unit of measure, such as tons
• Inventory “Quality,” including IQR and summaries of amounts of each type
• Customer service level, expressed how the CUSTOMER perceives it

ABC Analysis

Perform an ABC analysis, a simple, common and powerful tool for inventory management. It is based on Pareto’s law of “80-20.” The most common approach is to calculate demand in units, preferably for future periods, then calculate the total usage value at cost for each item (total cost of sales multiplied by units required) for a given future period. If future demand data are not available, the next best thing is to use history, but this won’t work well for items with major swings in demand over time. Sequence these in descending value. Typically, the top 10 to 15% of items account for 75-85% of value (“A” items), the next 20-30% account for 10-20% of value (“B” items) and everything else accounts for the rest, about 60-70% of the items, usually about 5% of the total value (“C” items). Your inventory should be less than these percentages for the “A” items, because they are much more tightly controlled and a little higher for B’s and significantly higher for C’s.

Then compare the list to actual values in inventory, plus actual and planned commitments. The answers will often suggest immediate corrective actions!

An ABC list suggests what to concentrate on to control most of the inventory investment. What it doesn’t tell you is that being short of a $.10 screw might prevent the shipment of a $5,000,000 radar unit, so ensure that there are control systems for all items, just control the expensive ones much more carefully. Err on the side of caution for the cheaper items, allowing a safety stock coverage or “two bin” approach to avoid stock outs, but keep inventory from getting out of control.

Create an Inventory Buildup Chart

Another good analysis tool is the inventory buildup chart. Use a standard x-y coordinate chart. Plot the cost build-up over time, by product group, with cost on the “y” (vertical axis) and time on the “x” (horizontal) axis. Normally, raw material cost accumulates first over time, followed by labor and overhead application. Allow for safety stocks, lot size inventory, transit stock, defects/rework/scrap, and normal finished goods and distribution pipeline stocking. Show the affect of consignment arrangements. Some people also treat accounts receivable as sort of a de facto inventory, until it is paid for. Once this chart is completed, show it around for shock value. Presented correctly, it will really make people think about the effect of constraints and decisions (just another form of constraint) on inventory. Then, work on changing the rules!

One company had a 14 month buildup curve, which was reduced to 4 months. At another company, the longest lead time material item accounted for only 20% of the product cost, so stocking only that item, instead of finished goods or instead of only reacting to orders, enabled them to radically reduce the response time for orders by 70%. It also added the flexibility of being able to use that raw material to make a number of different end items.

How to Identify and Control Inventory Drivers

Inventory drivers are things that tend to make inventory go up or down. Identify them and you will have some clue of why inventory changes. Understanding them is the beginning of gaining control. I’ve stated things that would drive inventory up, e.g.: more SKU’s. I refrain from stating the obvious: doing the opposite would reduce inventory. e.g.: reduce SKU’s to reduce inventory.

Key Drivers are covered briefly, as follows:

Number of SKUs

The more items you have, the more inventory you will need, in most cases. If you sell 500 widgets a year of A, then replace it with 250/year of A and 250 of B, you will probably need to carry more inventory. Why: demand and supply variability and total economic order quantities are likelier to be higher for 2 items than for one.

The more SKU’s in a product, the harder it is to bring matched sets of parts together at the same time. Because there are multiple items, with multiple vendors, kept and routed through multiple places or paths, with more opportunity for delays, defects, etc, more inventory will be needed.

The more operations there are and the longer that they take, the more inventory you will tend to have. More operations mean a longer supply chain. It may also mean differing lot sizes per operation and more places for delays and defects to occur. Process simplification helps reduce inventory.

The more facilities that inventory passes in and out of, the further apart those are and the harder they are to reach and pass material in and out of, the more inventory you will tend to have.

The more times inventory passes from the control of one system or organization to another and the less efficient the transfer is, the more inventory you will tend to have.

Lot/Batch Sizes

Lot/batch sizes greater than customer order delivery sizes tend to increase inventory. If customers order a product one at a time, but economics, handling or process considerations suggest that you make 1000 at a time, then you will have more inventory available than will be consumed per order, resulting in an accumulation of inventory. If you need to order things in cases, dozens, carloads, tons or weeks’ supply, but they are needed downstream in the supply chain in smaller increments, you will tend to accumulate more inventory.

The longer the lead time, the more inventory you tend to have. If something takes 16 weeks to get instead of 16 days, there is more inventory needed in process to cover the “pipeline” time. Whether it belongs to you or your vendor, it is increasing somebody’s cost, which ultimately will affect your cost and your customer’s cost. Longer lead time also means more chance of running out or having something go wrong out while waiting for it, which is usually dealt with by having additional inventory.

Carrying cost

This refers to the cost of owning inventory. For a closer look at these costs and the topic of inventory reduction, please see my article "Inventory Reduction - A How to Guide".

How to set Inventory Targets

After considering the current situation, drivers, and external situation, estimate what inventory levels should be, given certain sets of circumstances. There are impressive supply chain modeling tools to help you do this. Our experience is that developing an accurate detailed inventory behavior model is quite a chore to create and a major task to maintain, so we usually don’t. Normally working on projects with limited budgets, we study past behavior and focus on the main drivers, seeking to change a few with the greatest potential impact to achieve assigned objectives- sort of a “delta’ approach.

Don’t let us talk you out of sophisticated modeling tools, though. They have their place. When there are very large amounts of money involved and/or tricky constraints to work around, modeling tools will sometimes help. Many of the detailed control methods presented below contain elements of modeling.

Warning: Calculating or modeling inventory behavior solely by using the rules and parameters will nearly always be wrong. Why: If, for example, you assume that inventory will be an average of ? times the order quantity plus safety stock, you’ll most often be wrong. Actual supply and demand variability will differ. Defective items/customer returns may result in buildup. Unmatched sets of parts due to shortages will result in buildup. Generally, it is higher than the model would indicate.

Even the best laid plans can go off track if something changes unexpectedly- a major customer cuts orders, unexpected defects occur, requiring ad-hoc reaction, rather than careful, deliberate, advanced planning.

There are two major directions to approach inventory management from-- Top-Down and Bottom-Up. Most successful companies use a combination of both.

• Top-Down — this is the “macro” approach. Start with a goal, objectives, ABC (Pareto) analysis of estimated or historical usage, knowledge of overall processes and lead times. Set overall targets, by business unit at a minimum, preferably at a lower level, so that middle managers or even individual supervisors, work teams or administrative control personnel might be held more accountable. It takes more effort as the control is moved to a lower level.

Establish a tracking system, such as actual inventory versus target level. Compare numbers to actual sales, forecast. Monitor commitments and production plans against targets... Hold managers accountable for results and make them come back with reasons why targets cannot be met and solutions to the problems. Motivate them to solve underlying problems. Help them with problems outside of their scope of authority.

Another good tracking tool is Input-Output Control. Simply build a time-phased table of planned starting and ending inventories, showing starting, input, output and results. Then task employees to make the “delta’s” happen and track the actual values per period.

• Bottom-Up—Look at each item- determine cost, lead times, supply and demand reliability/variability, defect rate, transportation, storage, set-up/batch size considerations, buffers, process, handling considerations. Then set the proper planning methods and control parameters, to either default down from the enterprise, product line, commodity or department level to default down, or just establish them at the item/part level.

This takes a lot more effort than merely exercising Top-Down control, but it can deliver better results.

Educate and train people in inventory management and control approaches.

How to Control Inventory

After you do all your research and analysis, set targets and establish your control system, then you get to the hard part - actually making it happen.

Quick hits - Simply establishing the aggregate targets, understanding drivers, educating and training, setting up responsibility, establishing accountability and tracking results usually has significant effects. I have seen greater than 50% reductions from this alone. This can be the cheapest, fastest way of making some change happen, but it has a limited effect, because the approach lacks detail and won’t make major permanent changes in the ways that the business works without additional actions.

What is “Control?” - Control means to make something happen or to know why if it doesn’t, so that something might be done about it. Using that definition, there is no such thing as an uncontrollable situation. Someone once told me that he couldn’t control service inventory, because of unreliable vendor lead times. Nonsense! Unreliable lead times might be controlled by several strategies, such as: multiple sourcing, re-sourcing, safety stock, exhorting supplier to improve performance, ordering sooner, improving your own planning and reaction times, changing designs, alternate routing, training customers to order differently, having vendors stock raw materials. At least some of these would work in almost any situation.

Pitfalls of using control parameters

Wi

Here are typical Inventory System Metrics, which should be broken down by organization/responsibility, area, type, commodity, market/product, and time phased, with targets and actual values:

• Inventory Turnover or Days Coverage
• Inventory value or other unit of measure, such as tons
• Inventory “Quality,” including IQR and summaries of amounts of each type
• Customer service level, expressed how the CUSTOMER perceives it

ABC Analysis

Perform an ABC analysis, a simple, common and powerful tool for inventory management. It is based on Pareto’s law of “80-20.” The most common approach is to calculate demand in units, preferably for future periods, then calculate the total usage value at cost for each item (total cost of sales multiplied by units required) for a given future period. If future demand data are not available, the next best thing is to use history, but this won’t work well for items with major swings in demand over time. Sequence these in descending value. Typically, the top 10 to 15% of items account for 75-85% of value (“A” items), the next 20-30% account for 10-20% of value (“B” items) and everything else accounts for the rest, about 60-70% of the items, usually about 5% of the total value (“C” items). Your inventory should be less than these percentages for the “A” items, because they are much more tightly controlled and a little higher for B’s and significantly higher for C’s.

Then compare the list to actual values in inventory, plus actual and planned commitments. The answers will often suggest immediate corrective actions!

An ABC list suggests what to concentrate on to control most of the inventory investment. What it doesn’t tell you is that being short of a $.10 screw might prevent the shipment of a $5,000,000 radar unit, so ensure that there are control systems for all items, just control the expensive ones much more carefully. Err on the side of caution for the cheaper items, allowing a safety stock coverage or “two bin” approach to avoid stock outs, but keep inventory from getting out of control.

Create an Inventory Buildup Chart

Another good analysis tool is the inventory buildup chart. Use a standard x-y coordinate chart. Plot the cost build-up over time, by product group, with cost on the “y” (vertical axis) and time on the “x” (horizontal) axis. Normally, raw material cost accumulates first over time, followed by labor and overhead application. Allow for safety stocks, lot size inventory, transit stock, defects/rework/scrap, and normal finished goods and distribution pipeline stocking. Show the affect of consignment arrangements. Some people also treat accounts receivable as sort of a de facto inventory, until it is paid for. Once this chart is completed, show it around for shock value. Presented correctly, it will really make people think about the effect of constraints and decisions (just another form of constraint) on inventory. Then, work on changing the rules!

One company had a 14 month buildup curve, which was reduced to 4 months. At another company, the longest lead time material item accounted for only 20% of the product cost, so stocking only that item, instead of finished goods or instead of only reacting to orders, enabled them to radically reduce the response time for orders by 70%. It also added the flexibility of being able to use that raw material to make a number of different end items.

How to Identify and Control Inventory Drivers

Inventory drivers are things that tend to make inventory go up or down. Identify them and you will have some clue of why inventory changes. Understanding them is the beginning of gaining control. I’ve stated things that would drive inventory up, e.g.: more SKU’s. I refrain from stating the obvious: doing the opposite would reduce inventory. e.g.: reduce SKU’s to reduce inventory.

Key Drivers are covered briefly, as follows:

Number of SKUs

The more items you have, the more inventory you will need, in most cases. If you sell 500 widgets a year of A, then replace it with 250/year of A and 250 of B, you will probably need to carry more inventory. Why: demand and supply variability and total economic order quantities are likelier to be higher for 2 items than for one.

The more SKU’s in a product, the harder it is to bring matched sets of parts together at the same time. Because there are multiple items, with multiple vendors, kept and routed through multiple places or paths, with more opportunity for delays, defects, etc, more inventory will be needed.

The more operations there are and the longer that they take, the more inventory you will tend to have. More operations mean a longer supply chain. It may also mean differing lot sizes per operation and more places for delays and defects to occur. Process simplification helps reduce inventory.

The more facilities that inventory passes in and out of, the further apart those are and the harder they are to reach and pass material in and out of, the more inventory you will tend to have.

The more times inventory passes from the control of one system or organization to another and the less efficient the transfer is, the more inventory you will tend to have.

Lot/Batch Sizes

Lot/batch sizes greater than customer order delivery sizes tend to increase inventory. If customers order a product one at a time, but economics, handling or process considerations suggest that you make 1000 at a time, then you will have more inventory available than will be consumed per order, resulting in an accumulation of inventory. If you need to order things in cases, dozens, carloads, tons or weeks’ supply, but they are needed downstream in the supply chain in smaller increments, you will tend to accumulate more inventory.

The longer the lead time, the more inventory you tend to have. If something takes 16 weeks to get instead of 16 days, there is more inventory needed in process to cover the “pipeline” time. Whether it belongs to you or your vendor, it is increasing somebody’s cost, which ultimately will affect your cost and your customer’s cost. Longer lead time also means more chance of running out or having something go wrong out while waiting for it, which is usually dealt with by having additional inventory.

Carrying cost

This refers to the cost of owning inventory. For a closer look at these costs and the topic of inventory reduction, please see my article "Inventory Reduction - A How to Guide".

How to set Inventory Targets

After considering the current situation, drivers, and external situation, estimate what inventory levels should be, given certain sets of circumstances. There are impressive supply chain modeling tools to help you do this. Our experience is that developing an accurate detailed inventory behavior model is quite a chore to create and a major task to maintain, so we usually don’t. Normally working on projects with limited budgets, we study past behavior and focus on the main drivers, seeking to change a few with the greatest potential impact to achieve assigned objectives- sort of a “delta’ approach.

Don’t let us talk you out of sophisticated modeling tools, though. They have their place. When there are very large amounts of money involved and/or tricky constraints to work around, modeling tools will sometimes help. Many of the detailed control methods presented below contain elements of modeling.

Warning: Calculating or modeling inventory behavior solely by using the rules and parameters will nearly always be wrong. Why: If, for example, you assume that inventory will be an average of ? times the order quantity plus safety stock, you’ll most often be wrong. Actual supply and demand variability will differ. Defective items/customer returns may result in buildup. Unmatched sets of parts due to shortages will result in buildup. Generally, it is higher than the model would indicate.

Even the best laid plans can go off track if something changes unexpectedly- a major customer cuts orders, unexpected defects occur, requiring ad-hoc reaction, rather than careful, deliberate, advanced planning.

There are two major directions to approach inventory management from-- Top-Down and Bottom-Up. Most successful companies use a combination of both.

• Top-Down — this is the “macro” approach. Start with a goal, objectives, ABC (Pareto) analysis of estimated or historical usage, knowledge of overall processes and lead times. Set overall targets, by business unit at a minimum, preferably at a lower level, so that middle managers or even individual supervisors, work teams or administrative control personnel might be held more accountable. It takes more effort as the control is moved to a lower level.

Establish a tracking system, such as actual inventory versus target level. Compare numbers to actual sales, forecast. Monitor commitments and production plans against targets... Hold managers accountable for results and make them come back with reasons why targets cannot be met and solutions to the problems. Motivate them to solve underlying problems. Help them with problems outside of their scope of authority.

Another good tracking tool is Input-Output Control. Simply build a time-phased table of planned starting and ending inventories, showing starting, input, output and results. Then task employees to make the “delta’s” happen and track the actual values per period.

• Bottom-Up—Look at each item- determine cost, lead times, supply and demand reliability/variability, defect rate, transportation, storage, set-up/batch size considerations, buffers, process, handling considerations. Then set the proper planning methods and control parameters, to either default down from the enterprise, product line, commodity or department level to default down, or just establish them at the item/part level.

This takes a lot more effort than merely exercising Top-Down control, but it can deliver better results.

Educate and train people in inventory management and control approaches.

How to Control Inventory

After you do all your research and analysis, set targets and establish your control system, then you get to the hard part - actually making it happen.

Quick hits - Simply establishing the aggregate targets, understanding drivers, educating and training, setting up responsibility, establishing accountability and tracking results usually has significant effects. I have seen greater than 50% reductions from this alone. This can be the cheapest, fastest way of making some change happen, but it has a limited effect, because the approach lacks detail and won’t make major permanent changes in the ways that the business works without additional actions.

What is “Control?” - Control means to make something happen or to know why if it doesn’t, so that something might be done about it. Using that definition, there is no such thing as an uncontrollable situation. Someone once told me that he couldn’t control service inventory, because of unreliable vendor lead times. Nonsense! Unreliable lead times might be controlled by several strategies, such as: multiple sourcing, re-sourcing, safety stock, exhorting supplier to improve performance, ordering sooner, improving your own planning and reaction times, changing designs, alternate routing, training customers to order differently, having vendors stock raw materials. At least some of these would work in almost any situation.

Pitfalls of using control parameters

Wi

Key Drivers are covered briefly, as follows:

Number of SKUs

The more items you have, the more inventory you will need, in most cases. If you sell 500 widgets a year of A, then replace it with 250/year of A and 250 of B, you will probably need to carry more inventory. Why: demand and supply variability and total economic order quantities are likelier to be higher for 2 items than for one.

The more SKU’s in a product, the harder it is to bring matched sets of parts together at the same time. Because there are multiple items, with multiple vendors, kept and routed through multiple places or paths, with more opportunity for delays, defects, etc, more inventory will be needed.

The more operations there are and the longer that they take, the more inventory you will tend to have. More operations mean a longer supply chain. It may also mean differing lot sizes per operation and more places for delays and defects to occur. Process simplification helps reduce inventory.

The more facilities that inventory passes in and out of, the further apart those are and the harder they are to reach and pass material in and out of, the more inventory you will tend to have.

The more times inventory passes from the control of one system or organization to another and the less efficient the transfer is, the more inventory you will tend to have.

Lot/Batch Sizes

Lot/batch sizes greater than customer order delivery sizes tend to increase inventory. If customers order a product one at a time, but economics, handling or process considerations suggest that you make 1000 at a time, then you will have more inventory available than will be consumed per order, resulting in an accumulation of inventory. If you need to order things in cases, dozens, carloads, tons or weeks’ supply, but they are needed downstream in the supply chain in smaller increments, you will tend to accumulate more inventory.

The longer the lead time, the more inventory you tend to have. If something takes 16 weeks to get instead of 16 days, there is more inventory needed in process to cover the “pipeline” time. Whether it belongs to you or your vendor, it is increasing somebody’s cost, which ultimately will affect your cost and your customer’s cost. Longer lead time also means more chance of running out or having something go wrong out while waiting for it, which is usually dealt with by having additional inventory.

Carrying cost

This refers to the cost of owning inventory. For a closer look at these costs and the topic of inventory reduction, please see my article "Inventory Reduction - A How to Guide".

How to set Inventory Targets

After considering the current situation, drivers, and external situation, estimate what inventory levels should be, given certain sets of circumstances. There are impressive supply chain modeling tools to help you do this. Our experience is that developing an accurate detailed inventory behavior model is quite a chore to create and a major task to maintain, so we usually don’t. Normally working on projects with limited budgets, we study past behavior and focus on the main drivers, seeking to change a few with the greatest potential impact to achieve assigned objectives- sort of a “delta’ approach.

Don’t let us talk you out of sophisticated modeling tools, though. They have their place. When there are very large amounts of money involved and/or tricky constraints to work around, modeling tools will sometimes help. Many of the detailed control methods presented below contain elements of modeling.

Warning: Calculating or modeling inventory behavior solely by using the rules and parameters will nearly always be wrong. Why: If, for example, you assume that inventory will be an average of ? times the order quantity plus safety stock, you’ll most often be wrong. Actual supply and demand variability will differ. Defective items/customer returns may result in buildup. Unmatched sets of parts due to shortages will result in buildup. Generally, it is higher than the model would indicate.

Even the best laid plans can go off track if something changes unexpectedly- a major customer cuts orders, unexpected defects occur, requiring ad-hoc reaction, rather than careful, deliberate, advanced planning.

There are two major directions to approach inventory management from-- Top-Down and Bottom-Up. Most successful companies use a combination of both.

• Top-Down — this is the “macro” approach. Start with a goal, objectives, ABC (Pareto) analysis of estimated or historical usage, knowledge of overall processes and lead times. Set overall targets, by business unit at a minimum, preferably at a lower level, so that middle managers or even individual supervisors, work teams or administrative control personnel might be held more accountable. It takes more effort as the control is moved to a lower level.

Establish a tracking system, such as actual inventory versus target level. Compare numbers to actual sales, forecast. Monitor commitments and production plans against targets... Hold managers accountable for results and make them come back with reasons why targets cannot be met and solutions to the problems. Motivate them to solve underlying problems. Help them with problems outside of their scope of authority.

Another good tracking tool is Input-Output Control. Simply build a time-phased table of planned starting and ending inventories, showing starting, input, output and results. Then task employees to make the “delta’s” happen and track the actual values per period.

• Bottom-Up—Look at each item- determine cost, lead times, supply and demand reliability/variability, defect rate, transportation, storage, set-up/batch size considerations, buffers, process, handling considerations. Then set the proper planning methods and control parameters, to either default down from the enterprise, product line, commodity or department level to default down, or just establish them at the item/part level.

This takes a lot more effort than merely exercising Top-Down control, but it can deliver better results.

Educate and train people in inventory management and control approaches.

How to Control Inventory

After you do all your research and analysis, set targets and establish your control system, then you get to the hard part - actually making it happen.

Quick hits - Simply establishing the aggregate targets, understanding drivers, educating and training, setting up responsibility, establishing accountability and tracking results usually has significant effects. I have seen greater than 50% reductions from this alone. This can be the cheapest, fastest way of making some change happen, but it has a limited effect, because the approach lacks detail and won’t make major permanent changes in the ways that the business works without additional actions.

What is “Control?” - Control means to make something happen or to know why if it doesn’t, so that something might be done about it. Using that definition, there is no such thing as an uncontrollable situation. Someone once told me that he couldn’t control service inventory, because of unreliable vendor lead times. Nonsense! Unreliable lead times might be controlled by several strategies, such as: multiple sourcing, re-sourcing, safety stock, exhorting supplier to improve performance, ordering sooner, improving your own planning and reaction times, changing designs, alternate routing, training customers to order differently, having vendors stock raw materials. At least some of these would work in almost any situation.

Pitfalls of using control parameters

Wi

Even the best laid plans can go off track if something changes unexpectedly- a major customer cuts orders, unexpected defects occur, requiring ad-hoc reaction, rather than careful, deliberate, advanced planning.

There are two major directions to approach inventory management from-- Top-Down and Bottom-Up. Most successful companies use a combination of both.

• Top-Down — this is the “macro” approach. Start with a goal, objectives, ABC (Pareto) analysis of estimated or historical usage, knowledge of overall processes and lead times. Set overall targets, by business unit at a minimum, preferably at a lower level, so that middle managers or even individual supervisors, work teams or administrative control personnel might be held more accountable. It takes more effort as the control is moved to a lower level.

Establish a tracking system, such as actual inventory versus target level. Compare numbers to actual sales, forecast. Monitor commitments and production plans against targets... Hold managers accountable for results and make them come back with reasons why targets cannot be met and solutions to the problems. Motivate them to solve underlying problems. Help them with problems outside of their scope of authority.

Another good tracking tool is Input-Output Control. Simply build a time-phased table of planned starting and ending inventories, showing starting, input, output and results. Then task employees to make the “delta’s” happen and track the actual values per period.

• Bottom-Up—Look at each item- determine cost, lead times, supply and demand reliability/variability, defect rate, transportation, storage, set-up/batch size considerations, buffers, process, handling considerations. Then set the proper planning methods and control parameters, to either default down from the enterprise, product line, commodity or department level to default down, or just establish them at the item/part level.

This takes a lot more effort than merely exercising Top-Down control, but it can deliver better results.

Educate and train people in inventory management and control approaches.

How to Control Inventory

After you do all your research and analysis, set targets and establish your control system, then you get to the hard part - actually making it happen.

Quick hits - Simply establishing the aggregate targets, understanding drivers, educating and training, setting up responsibility, establishing accountability and tracking results usually has significant effects. I have seen greater than 50% reductions from this alone. This can be the cheapest, fastest way of making some change happen, but it has a limited effect, because the approach lacks detail and won’t make major permanent changes in the ways that the business works without additional actions.

What is “Control?” - Control means to make something happen or to know why if it doesn’t, so that something might be done about it. Using that definition, there is no such thing as an uncontrollable situation. Someone once told me that he couldn’t control service inventory, because of unreliable vendor lead times. Nonsense! Unreliable lead times might be controlled by several strategies, such as: multiple sourcing, re-sourcing, safety stock, exhorting supplier to improve performance, ordering sooner, improving your own planning and reaction times, changing designs, alternate routing, training customers to order differently, having vendors stock raw materials. At least some of these would work in almost any situation.

Pitfalls of using control parameters

With the use of MRP, MRPII, ERP and now “Supply Chain Management “ systems, there are more opportunities to improve inventory management, but also more chances to lose control! Unless there is a clearly stated Aggregate Inventory Management approach imbedded in the system, through education, training and parameters, yes- I said parameters!, you will likely fail.

War story from George Miller: “After I left a specialty niche MRPII/ERP company for the consulting world, a customer of that company called to inform me that the “software wasn’t working.” The problem was that the system was carrying out their instructions at the speed of light, spewing forth recommendations to acquire inventory, based on their unrealistic parameters. Most of these systems have various ‘gauges’ and “levers,” to set control parameters to tailor the operation of the system to the company, products and process. These might be set, for example, system-wide, but can usually be overridden at the business unit, plant, department, product line and/or part number level. Each level normally defaults down to the lower level, unless you override it.

“For example, they used unrealistically long process times in the item master planning records and had safety stock and scrap factors planned at multiple levels in the bill of material, “pyramiding” (increasing) demand calculations considerably. No surprise then, except to them, that they were well upon their way to doubling their inventory investment in record time, without significant benefits. The prescription was:

1. The management team to get personally involved in setting the system parameters.

2. Educate employees in inventory management concepts and train them in proper use of system tools.

3. Establish and monitor a special report to assess the effect of “order modifier” parameters, such as safety stock, scrap and attrition factors, order planning method, order quantity rules, order multiples, lead time, review time, inspection time.”

Conclusion

Inventory can be systematically managed. It doesn’t happen on its own. Needed is a rationale, a plan, education, training, organization, tools, policies, procedures and management willpower.

This article is also available in full on our website: PROACTION – Generating Best Practices. It is an excerpt of a paper originally written by George Miller, Founder of PROACTION. It has been modified and updated by Paul Deis, PROACTION CEO.