|
Parts Inventory Counts |
.. Feb 2/05, Geoff |
|---|
|
Related Pages |
There are 2 types of inventory counts: "periodic" counts, where we count
everything once a year; and "perpetual" counts, where we count a little bit
each week.
The main advantage of the perpetual inventory count model
that I based the Foxpro > Count program on, is that we can fine tune the
count to target those items which are most likely to need counting.
Here is the logic behind the Foxpro > Count program:
- right now,
about 94% of the parts on our shelf match what is in the computer, and 6% are
wrong. (This figure is based on Frontier's historical averages.)
- in
Delta, that means that we probably have 800 wrong items on the shelf; Edmonton
probably has 500; Calgary probably has 320
- wrong quantities are a
problem: the parts department spends extra time chasing down missing parts;
customers don't get what they want when they want it; production and shop
repairs get stalled; we order things we don't need to order; we don't know for
sure what components we used to build our equipment (which creates problems
when a customer wants a replacement piece); and our accounting values don't
match reality
We could count every item in the warehouse every
month. Surprisingly, this is not the best plan, because the counting process
itself creates mistakes. Some of these problems are caused by parts on reserve
(are they still on the shelf or in a box somewhere?); in-process interbranch
transfers; parts that may have been received but not yet put into the computer;
vendor invoicing problems & parts receiving problems (we have about 100 of
these every month); and inventory-counter fatigue.
For example, the
one-shot complete inventory counts in Edmonton in previous years have turned up
hundreds of adjustments that then get un-done in subsequent months.
The task is to find the wrong quantities with the least amount of
effort.
- Instead of counting every part on the shelf, we can find
most of the problem items quickly and easily by targeting the parts that most
need to be counted.
The Foxpro Count program prioritizes items based
on 3 factors:
- the parts movement code (A,B,C,D...) This is calculated on
the gross margin contributed by that part number in the past year. This parts
movement code is recalculated every night. We don't need to count parts that
haven't sold in several years very often; fastest-moving parts are more
important to count.
- the date this part was last counted
- the
quantity on the shelf (e.g. it's easier to count if there are between 1 and 5
on the shelf.)
- We temporarily ignore items where the conditions might
create problems (parts on order, on reserve, or in-transit) and count them the
following week.
Using these criteria, the Count program generates
a list of parts that most need to be counted.
- A typical Count list
in Delta finds 15% of the parts with wrong quantities. Finding 15% of the
target group vs. 6% of all parts with wrong quantities is 2.5 times more
efficient. More efficient means less work. Who doesn't like that?
At the moment we are behind on our inventory counts.
Once we are
caught up, we only need to count about 4% of our inventory each month. That
works out to 49 part numbers a week in Calgary; 72 part numbers a week in
Edmonton; and 119 part numbers a week in Delta. If we split that among all the
parts department employees, it's not that much work to keep our inventories
accurate, and our customers, shop, management and accountants happy.
In fact, we will probably spend LESS time counting inventory that we currently
spend compensating for inaccurate counts (by scrambling to get something in
that we thought we had on the shelf; or by having to double-count items because
we don't believe what's in the computer.)