|
Managing
On the Other Side
by Dave Garwood
Visualize
the supply chain for any goods or service. Demand starts at the top with
the ultimate consumer and drives down to the supplier of raw material,
along the way passing through retailers, distributors, manufacturers and
suppliers. Supply, on the other hand, starts at the bottom, increasing
value at each step, before reaching the consumer.
The
focus of Supply Chain Management is to simultaneously reduce total costs
in the entire chain and make the whole process more responsive to customers'
needs. Making supply equal demand at every stage in the chain is critical.
Where,
though, has our primary focus been? In most cases, on the supply side of
the equation. The old paradigm was, "We manage supply, but demand just
happens." Demand has remained a wild card. We've traditionally accepted
wild demand variations as a way of life because demand comes from customers
and, "we can't control our customers." The new paradigm is we must manage
both supply and demand to have an effective, competitive supply chain.
Before
you zone out, this isn't an article about getting an accurate sales forecast.
This is an article about managing the variation between planned and actual
demand. When actual demand is compared to the plan, variations are common.
Our goal is to make supply equal to demand. This means we want to make
what the customer wants, deliver it when the customer wants it, sell it
at a price the customer is comfortable paying and still make a buck! If
we have that likely demand variation, what are the alternatives to keep
the equation equal?
1.
Build inventory to buffer the variations. This works fine as long as we
are willing to have large sums of money tied up in products sitting in
a warehouse somewhere. It also works only if you inventory the right items
-- the items the customers eventually want to buy. Realistically, that
often doesn't happen. If we knew what the customer wanted, we wouldn't
need to build inventory in the first place. And the hidden costs of moving,
storing, tracking, etc. are huge!
2.
Vary the customer lead times. Explain to customers that they ordered at
the wrong time! Make them wait until the product is available. This may
be an open window of opportunity for your competitors.
3.
Flex the supply up or down every time demand goes up or down. This may
work sometimes, but it can be very costly or impossible to flex the supply
enough if the variations are significant.
4.
Eliminate the demand variation. Teach the customers to read our forecast.
Right! Demand variation may be reduced, but eliminated is a pipe dream!
All
of these alternatives have significant downside consequences. The best
strategy is a fifth alternative --simultaneously minimize demand variation
and develop the ability to flex the supply within reasonable limits. Now
let's look at minimizing the demand variation.
Reducing
Demand Variability: A Quality Issue
Think
of demand planning as a "process." The quality revolution left some valuable
lessons to help minimize variation in any process -- including demand planning.
Lesson
1: Defects, defined as variation outside acceptable limits, are not
inevitable. Plan to have defects, and you'll get them! An essential paradigm
shift here is that both actual supply and demand can be managed, i.e.,
brought within a reasonable expectation of the plan.

Lesson
2: Every process has variability; therefore, expect actual sales to
be above and below the plan. What is an accurate forecast? If given truth
serum, I suspect we would say actual sales equal the forecast. It will
never happen. No wonder sales people often run for cover when asked to
participate in the forecasting process, especially when we measure forecast
"accuracy."
Lesson
3: Variation is costly! Reduced demand variation translates into higher
customer satisfaction, which means greater profits. It also means sales
has more time to sell and spends less time doing damage control.
Lesson
4: A proven process exists to eliminate defects. First, establish a
plan, the desired outcome; in this case, the demand plan. Then, measure
actual performance; in this case, actual sales. Establish acceptable tolerances;
in this case, an anticipated range around the demand plan. If actual demand
is not within the range, determine the root causes for the excessive variation,
brainstorm solutions and assign responsibility for corrective actions.
Once
demand is under control, tighten the "tolerances." Continuous quality improvement,
not accuracy, is the goal. Quality of the plan is defined as the number
of items where actual demand falls inside the tolerances, not right on
the plan. Finding root causes and taking preventive action is the key to
getting demand plans under control. Most companies tell us that demand
is uncontrollable, because customers are uncontrollable. But in almost
every case we examine, many of the wild swings in demand are, in fact,
self-inflicted wounds. Often, it's a communications weakness in the supply
chain. Sales promotions, quotes or new marketplace activities don't get
communicated. Lot sizing can send a cascading message down the chain. The
consumer buys one, the store orders a dozen, distribution orders a gross
and the factory produces an "economical" truckload quantity. And then no
more orders for several months!
Other
examples of root causes include financial pressure for end-of-the-quarter
sales pulls demand from future months into the current quarter or multiple
forecasts. Manufacturing has their forecast, which doesn't even resemble
the plan from Sales & Marketing. And Finance doesn't use either plan. Which
one do we measure and manage? These steps worked to eliminate defects in
the factory and will work to minimize demand variation as well.

Can
we bring the demand process down to zero variation? Of course not. We can,
however, get the demand plan under control, that is, a predictable variation.
Many of the companies we work with report excellent immediate results by
virtue of the fact that, for the first time, they are looking at demand
planning as a process, rather than a force of nature. Just by eliminating
our self-inflicted wounds, we take many of the wild swings out of the demand
process. Remember, "defects" are not inevitable, even in demand planning!
|