The Inventory Trade-off

Inventory is a necessary evil in modern supply chain. Identifying and
maintaining the right amount of inventory is one of the biggest challenges that
supply chain managers face. Inventory sits as a trade-off between customer
satisfaction and material availability. With globalization of organizations,
Indian companies need to match their international peers in terms of
sophistication and maturity of supply chains. However, maturity of supply chains
and the supporting technology does not completely eliminate the aforesaid
trade-off completely.

IBM, in partnership with IW Custom Research, undertook a value chain survey,
the results of which were published in early 2008. IBM also conducted a Value
Chain Survey in India, in partnership with Economic Times Intelligence Group,
which was published in 2007. This article draws from the reports and IBMs vast
experience in handling inventory-related strategic and operational projects
across industries.

The article highlights the top-push seven challenges that are critical but
not obvious, to retain a balanced inventory in the supply chain, preventing
unproductive working capital, and lost sales opportunities.

1. Infrequent Parameter Reviews

The parameters used for managing inventory, such as safety stock quantity,
replenishment order quantity, reorder point in a continuous review policy, or
review period in a periodic review policy, use factors such as service levels,
demands, and supplier replenishment lead times as inputs for their calculation.
However, rapidly changing markets, competitors, and product life cycles have
made review periods that worked in calmer times unsuitable for todays speed of
business execution. Failure to monitor the environment and update these inputs
on a frequent and detailed basis is a recipe for inefficient inventory
investment. The untracked accumulation of excess inventory leads to inventory
write-downs and lost working capital.

Solution: Periodic review of key planning parameters is required for ensuring
proper inventory management. Tools and applications are effective enablers of
such reviews.

2. Days of Supply

When faced with unpredictable demand and pressures to ensure supply
availability, a seemingly logical approach is to declare a quantity of supply to
be held at all times for each SKU. Typically, this quantity is expressed in
terms of days of supply, or days coverage. The approach reasons that if it
takes, on an average, two weeks to receive replenishment from when the order is
placed, then adding a safety cushion and calculating the average daily
consumption will be fine.

Many inventory managers concede that by taking this approach, they may be
giving up inventory reduction (and cost saving) opportunities in order to
guarantee availability. However, this approach isnt completely suitable to
determine an efficient supply quantity. It fails in three areas: first, it does
not take into account the daily, seasonal, and lifecycle variations in demand
that may be inherent for each SKU; second, it does not take into account
variations in supplier replenishment lead time; and third, it does not seek a
cost-efficient balance between inventory held as safety stock and inventory
being replenished (cycle stock).

Solution: Monitor and understand the variations in demand and supplier lead
times, and frequently recalculate safety and cycle stock for each stock-keeping
unit.

3. Volume-based Classification

With tens and, in some cases, hundreds of SKUs to manage, some
classification and prioritization is a required first step. However,
classification based only on quantity sold is not the most optimal way. What may
seem intuitive may not be the most effective approach for managing inventory at
the lowest overall cost. With the advent of ERP systems and other
high-computation enablers, it is now possible for a more fine-grained approach
to material classification beyond simple methods like FSN.

Solution: Consider alternative classification techniques that incorporate
multiple costs and parameters to optimize inventory investment.

4. Top-down Forecasting

In many cases, the method for generating the forecast is set by company
division or product type. Thus, a single method of forecasting is cascaded down
to product lines and perhaps even to the individual SKU/location level. Although
an aggregate forecast approach is necessary for manufacturing and procurement
capacity planning, and for supplier negotiations, it is not the most effective
means for optimizing the investment in inventory.

Such top-down approaches do not account for the local variations in demand
that can be significant at the individual SKU level. Even within a single
product type, some SKUs are relatively new in their lifecycle, others are soon
to be retired; some are affected by seasonality, others are not, but may be very
sporadic in both time and relative quantity between sales. Solution: Create
forecasts at the SKU stocking level to generate reliable models of demand
variation.

5. Misaligned Service Level Measurement

The service level target is a key factor in the determination of inventory
levels. Yet, despite its influence, rarely enough attention is paid to the
calculation of this measurement.

Implicit in most service level discussions is that poor service levels result
in lost sales and vice versa. Therefore, an effective service level measurement
should reflect customers performance expectations and buying behavior.

In addition, understanding the link between service level and customer
satisfaction also opens the door to segmenting the customer base to further
differentiate service levels by customer segment.

Solution: Create an approach to service level calculation that can provide
multiple differentiated measurements of serviceability and include all functions
in determining which measurements and values should be used when developing
inventory plans.

6. Addressing the Causes

When science is applied to understanding service levels, supply variability,
demand variability, and various stocking policies, the result, though
technically optimal, still holds opportunity for improvement. This is because
variability in both supply and demand can often be self inflicted or turn out to
be easily addressable.

The right inventory level for a given level of service is highly sensitive to
the variability around supply and demand.

Solution: Understand the sources of high variability on both the demand and
supply sides of your stock, and take actions to reduce the variability.

7. Failure to Align Inventory Policies

Reduced inventory helps in greater working capital productivity. However,
resistance from sales and marketing, due to the risk to service levels, is high.
An approach that simulates various inventory policies and parameters can bring
transparency in the decisions and their impacts.

Simulations allow companies to compare the impacts of different parameter
choices on expected stock-outs, service level, and inventory costs. Also, by
simulating under different demand and supplier response scenarios, a firm may be
able to anticipate the correct level of inventory to hold, to minimize its
exposure to risk. The ability to see how a companys inventory will fluctuate
next week and in the months to come is imperative for reducing the fear around
change, and helps to align the organization on priorities.

Solution: Invest effort in building organizational alignment around inventory
policies, potentially using simulation capabilities to anticipate the expected
results of your inventory parameter decisions. Also, especially in a
geographically dispersed country like India, carefully examine the distribution
structure and its impact on inventory levels and analyse distribution routes
based on the velocity of the product.

Over the years, inventory management has become a powerful tool to deliver
superior services efficiently. Excellence in inventory management requires not
only a one-time assessment of the issues facing the organisations supply chain,
but a regular alignment to suit changes both internal and external. Rule of
Thumb decisions need to be replaced by an approach that blends data analytics,
optimization algorithms, and discrete event simulation.

Clifford Patrao & Suresh Chivukula

maildqindia@cybermedia.co.in

Clifford Patrao is an associate partner, IBM India and is the lead, Consulting
Services division.Suresh Chivukula is a consultant in the supply chain area.