THIS IS AN INTERNAL GOLDRATT GROUP DOCUMENT.

IT IS NOT INTENDED FOR GENERAL DISTRIBUTION

Never say, I know! Page 1 Eliyahu M. Goldratt, 2006


Never say, I know!

Eliyahu M. Goldratt, 2006


If there is a rule that Mickey and I try to drill into everyone who is doing
two-hour meetings, it is: check and cross-check the key data on which
our suggested solution is based. Last month I had to face a case where
it was apparent that I had deviated from that rule. Of course the
consequences were embarrassing, but that is not the reason Im now
forcing myself to sit down and write this document (Im not a
masochist). The main reason is that the renewed analysis showed me
(again) the extent to which there is no end to deeper understanding;
we have embarked on a journey that does not have an end, just
exciting and rewarding stepping stones.

Maybe I was hit by this case because the company is producing
sporting apparel. If there is an industry that I fooled myself that I know
inside and out, it is the apparel industry. Nevertheless, the renewed
analysis yielded not one but three additions to the body of knowledge.
The importance of these additions can be evaluated by the fact that in
the past few weeks I have already used them effectively in two other
cases.

Here are the details. Eighty-five percent (85%) of the companys sales
is to brand companies (companies like BigBrand which was the
subject of the analysis described in a previous document). But unlike
the cases we see in China and other places, this company has over 10
brand companies as clients, none of them dominating its sales.
Therefore I wasnt overly surprised to read that raw material out of the
selling price is only 50%. (When you are in the hands of one dominate
client expect your prices to be squeezed and therefore expect that
materials will represent a higher percent of your sales.) The other 15%
of sales are their own collection, their own brand, that they sell through
their own 10 shops and 4 franchised shops, all located in their small
country.

A text-book case. In the two-hour meeting I just verified that their
production lead time is two months (very common in the apparel
industry) and yes, they produce for the entire season in one batch. I
also verified that every client has its own collection and as a result there
is no aggregation. The template was obvious; the new awakening of the
THIS IS AN INTERNAL GOLDRATT GROUP DOCUMENT. IT IS NOT INTENDED FOR GENERAL DISTRIBUTION
Never say, I know! Page 2 Eliyahu M. Goldratt, 2006

brand companies will guarantee that they will welcome a VMI offer. No
real risk here since it will be enough to convince just 3 or 4 of the
existing (huge) clients and we can increase sales as much as we want.
Of course it will take one or two seasons for the clients to realize the
amazing benefits a VMI contractor brings to them and as a result to get
much more work.

There is no real set-up in production of sports apparel, so it should be
easy to cut the lead time to less than a week. Transportation time to
the central warehouses of all their clients is only few days. It should be
a walk in the park to bring them to excellent performance on a VMI
offer.

Increasing capacity is easy; there is no shortage of inexpensive labor to
sew the goods and the machines are just sewing machines. At 50%
material out of selling price the VV seems relatively easy to reach,
especially when we have four years. Next case please.

During the 4x4 it turned out that a key element was wrong raw-
material is not 50% of sales, rather it is 75%. This, as you know,
changes the entire picture. Lets substantiate this last sentence. When
RM is 50%, to reach a VV, we have to triple the volume. Baring in
mind, the existing access capacity and the substantial improvements we
bring in logistic (it is an A plant), the required increase in volume
translates into a manageable increase in just direct labor (~ 50% or
less). But when RM is 75%, to reach a VV, we need to multiply the
volume by a factor of 5. Under the same realistic assumptions, that will
necessitate increasing direct labor by about 150%. Such major
increase will require increase in supervision, administration etc. That is
no longer negligible in its impact on the company OE. To compensate
for the increase in OE, and since the margins are small to start with, we
need to increase sales much more. The downward spiral is evident.
The only way to get out of this spiral is to find a way, not just to
increase sales, but to significantly increase margins as well. In short,
kiss good-bye a VV which is based on straight VMI.

Two different questions should be answered. To avoid running again
into such a dead-end we need to know: how come such a basic piece of
data was wrong? To ensure that we do have VV for contractors to
brands we should also try and answer a more interesting question: is
there a feasible way to increase margins?

As for the first question, it didnt take long to reveal the source of the
mistake. The number was derived from their financial statement and
therefore it represents an average over the two channels of sales.
THIS IS AN INTERNAL GOLDRATT GROUP DOCUMENT. IT IS NOT INTENDED FOR GENERAL DISTRIBUTION
Never say, I know! Page 3 Eliyahu M. Goldratt, 2006

Even though the other channel direct sales through their own shops
is relatively small (just 15%) it had a major impact on the average
percent of RM. Simply, for the channel of direct sales, the TVC
represents much, much less than 50% (margins in the direct sales
channel are composed of the huge markup of a brand plus the
substantial markup of a shop). The other distortion was due to the fact
that they pay by piece for the people who sew, turning most direct labor
into (by TOC definition) raw-materials.

We now understand the source of the mistake. But it doesnt help us to
answer the much more interesting question: what can we do to increase
margin?

Moving to rapid response is the first thing that jumps to mind,
especially when one considers the current production lead-time of two
months and the ease to cut the production lead time to be, at most, one
week. Well, in their environment there is a limitation that makes it
difficult. Dying the fabric is a batch process. You need much more than
willingness to pay slightly higher prices to convince a supplier to cut all
the batches to be in line with one week consumption (cutting the
batches to be less than one tenth of what they are now). Such a
change will necessitate a super culture revolution. Moreover the
supplier is not willing to guarantee the same color in subsequent
batches.

Knowing the amazing benefits that moving from producing to forecast
to producing for replenishment brings the brand companies, I first
concentrated on finding a way to solve the problem of large batches of
dyed fabric. Instead of buying dyed fabric in small quantities, maybe
we can still buy it in the same large quantities and, relying on the fact
that the same dyed fabric is used for more than one SKU, we can use
the fabric according to actual demand.

The extent to which we can do this depends on the magnitude of the
aggregation that exists at the dyed fabric level. We know that there is
aggregation since all sizes of the same model use exactly the same
dyed fabric. Unfortunately, in sporting apparel there are only five sizes
per model (as compared to over 30 sizes in formal suit jackets).
Moreover, there is a high chance that if one size is a high-runner, the
other sizes are not slow movers; there is a correlation between the
demand for one size and the demand for another size of the same
model. Therefore, the aggregation that we get through different sizes
will not help us much there is no point to divert dyed fabric to one
size just to find, a month later, that we cannot provide another needed
size.
THIS IS AN INTERNAL GOLDRATT GROUP DOCUMENT. IT IS NOT INTENDED FOR GENERAL DISTRIBUTION
Never say, I know! Page 4 Eliyahu M. Goldratt, 2006


So, it all depends on the commonality of the dyed fabric (magnitude of
aggregation) between different models where there is almost no
correlation between the demand for one model and the demand for
another. For that I checked the number of different models they
produce per year (35,000) and compared it to the number of different
dyed fabrics they use per year (4700). The ratio is one to seven. Is it
enough?

We know that, in general, about 30% of the SKUs are depleted before
the massive end-of-the-season clearance sale starts the high runners.
Another 30% are sold mainly in the end-of-the-season sales (and in the
outlets) the slow movers. That means that in a group of seven
models there is a high chance that at least one model will be a high
runner while at least another model will be a slow mover. So even
though we start with a given (forecasted) amount of dyed fabric, still, in
most cases, the magnitude of diversion that we can do (from the actual
slow movers to the actual high runners) will help. How much will it
help?

To answer this question we have to realize the non-linear nature of the
damage shortages are causing. Suppose that a particular SKU has been
depleted after one month and the season is four months. How much
are the lost sales relative to the amount that was sold? One doesnt
have to be a trained mathematician to answer this question; the lost
sales are 300% relative to the actual sales. What about an SKU that is
completely sold out after three months? The lost sales are just 33%
relative to the actual sales (less, if one takes into account that in the
last month of the season prices are lower). That means that even
though we work with restricted availability of the colored fabric we do
get most of the effect, in terms of percent increase in sales, since most
lost sales are prevented. The same goes for reducing the bad effect of
obsolescence, especially when we consider that the longer it takes until
the slow mover is sold the lower the price at which it is sold.

How can we take advantage of it? How can we turn it into a mafia offer
to the brand companies? Currently the brand companies order the full
amount per SKU and demand delivery before the beginning of the
season. Once they get the goods they immediately push about 40%
into the retail to fill up the pipelines with the new collection. Lets not
try to change these big companies. Lets give them an offer that does
not require any real change from their side and that the advantages of
what we offer will be apparent to them.

THIS IS AN INTERNAL GOLDRATT GROUP DOCUMENT. IT IS NOT INTENDED FOR GENERAL DISTRIBUTION
Never say, I know! Page 5 Eliyahu M. Goldratt, 2006

What do you think about the following offer? As it is currently done
well get the orders per SKU (based on forecast) enough time before the
season. Based on the orders well buy the entire quantity of dyed
fabric, as we currently do. But, we will cut, sew and ship only half the
quantity of each SKU. Now we wait for the orders to come to the brand
company from the retailers. The first orders are apparently orders for
high runners. We are informed of each order that the brand gets, and,
as long as we still have that colored fabric, we will replenish within a
time period which is very small relative to the period of the season (two
weeks should be enough for a system that is used to minimum two
months). Six weeks before the end of the season, the brand should
instruct us what we should do with the residual fabric. Should we turn
it into goods or should we hold it for the next year (on the brands
expense)? I think that if such an offer is presented well (in the SFS
way) the chances are very high that each and every brand will gladly
accept.

As a matter of fact, knowing that all brands are currently struggling to
find ways to increase their inventory turns I think that they will be
interested in such an offer to the extent that we can use it effectively to
increase our margins. Here Im going into the speculative mode; the
following should be checked with the brands and modified according to
their response.

Brands that operate on three seasons a year have about 6 inventory
turns. For them, increasing to nine inventory turns is a major
achievement, and I doubt that anyone in those companies thinks that
twelve inventory turns is realistic for his company. Therefore,
requesting a bonus based on actual inventory turns might fly,
something along the lines of: right now, in sporting apparel you have
6 inventory turns. With our offer, which is based on the mammoth
effort that we have done to reduce our reaction time from over two
months to less than two weeks (including transportation), we think you
can enjoy higher inventory turns. Lets take the conservative
assumption that fluctuations (and a lot of luck) might bring your
inventory turns on our goods to eight. So lets acknowledge our
contribution to your improved results only if inventory turns go up to
the delightful number of 9. But then compensate us for our unique
contribution. For example, for each inventory turn starting at 9 give us
a bonus equal to just 5% of our price.

Since the mark-up of the brands is usually around 400% of the price
they pay to their contractors and since increasing their inventory turns
is of such importance to them, such an offer has a real chance of being
accepted (at least by 3 or 4 of the 12 clients the company has). Of
THIS IS AN INTERNAL GOLDRATT GROUP DOCUMENT. IT IS NOT INTENDED FOR GENERAL DISTRIBUTION
Never say, I know! Page 6 Eliyahu M. Goldratt, 2006

course to conclude such a deal well have to climb the ladder, from the
purchasing person to a relatively high-level manager, so dont expect a
quick deal.

What impact will it have on our margins? My founded expectation is
that such an offer will actually bring the inventory turns of a brand
(without any action from the brand to change the way it does business
with retail) from 6 to 15* turns. The offer has the potential to double
the companys margins while increasing dramatically its sales.


[*At first glance 15 inventory turns seems optimistic but, actually, it is
very conservative. Just consider that right after the beginning of a
season a brand currently holds 60% (100%-40%) of the goods (this
number is going down slowly during the season) while using our way
the brand will hold just 10% of the goods. So we can expect the
inventory turns will increase five fold and that is without taking into
account that sales will go up 15 inventory turns is very conservative.]

You have probably noticed that I have some slight hesitations. Simply
we havent yet tried getting bonuses for improved inventory turns, or
anything similar. So I kept on thinking of additional ways (to be done
in parallel) to increase the companys margin while increasing
dramatically their sales.

How to get higher margins? I couldnt think of an additional way to get
a higher price from the companys clients - the brand companies. So
what about bypassing them and selling directly to the retail stores?

Usually, this avenue is not really feasible for a contractor since it
actually demands building a new type of a company. It is one thing to
have the ability to cut and sew fabric into garments according to given
designs, it is a totally different ability to design new collections of
garments and to do it three times a year in pace with the changing
fashion. But in this particular case our company already has this ability.
They do have their own collections and in their country their collections
compete well with other brands collections. As a matter of fact, in their
country they are number three in selling their own sports apparel,
ahead of many, much better known and much bigger brands. So if we
want to increase the companys margin, we just have to concentrate
more on sales directly to the shops. And since they have almost
saturated their small country - they have their shops in every major
location - they will have to approach shops outside their country.
Considering the markup of brands, this will not just increase the
companys margins, it will explode them.
THIS IS AN INTERNAL GOLDRATT GROUP DOCUMENT. IT IS NOT INTENDED FOR GENERAL DISTRIBUTION
Never say, I know! Page 7 Eliyahu M. Goldratt, 2006


Isnt it obvious? Well, from harsh experience, Ive learned to be very
careful in answering this question. On the one hand I do recognize that
all good solutions have one thing in common, they are obvious but only
in hindsight. Always, once I finally verbalize a good solution to a major
problem, Im disappointed with myself for wasting so much time until I
reached the obvious. But, on the other hand, I also learned not just to
respect, but actually to admire, peoples experience and intuition. If the
solution is correct, if the solution is really so obvious, how come that
people havent used this solution a long time ago? It must be that
something, an erroneous assumption that they took for granted as an
undisputable fact of life, caused them to dismiss this solution; blocked
them from even trying to implement it. So until I clearly recognize and
verify such a blocking assumption I dont know if my new solution is
obvious, or just stupidly wrong.

Why hasnt this company tried to sell directly to shops outside its own
country? It must be because they have the name recognition, the
brand name, mainly within the borders of their own country (in other
words, outside their country they are a non-brand company). And the
managers of the company had learned that to build a brand-name takes
time, a lot of time, and money, a lot of money. The bigger the country
the bigger the sums of money required. It is obvious that to build a
brand name in a sizable country is beyond their current financial (and
managerial) abilities.

But why is having a brand name so important? Why are they convinced
that as long as the company hasnt yet established a real brand name,
in the territory in which a shop is located, it will not be economical to
try and persuade the shop to carry the companys collection?

It is, probably, because shops know that merchandise that carries a
known name will sell and they are reluctant to take the risk of buying
non-brand merchandise, merchandise that might not sell well enough.
The shops' reluctance is logical because the constraint of most shops is
display space (and cash). Carrying merchandise that might not sell well
is actually wasting the constraint (the opposite of exploit) and
therefore reduces the throughput of the shop.

We can proceed from here in two ways. One is systematic, logical and
meticulous. The other is bold, aggressive and not less logical. (Actually
there are many ways to proceed, but since they are not logical Ill
ignore them.)

THIS IS AN INTERNAL GOLDRATT GROUP DOCUMENT. IT IS NOT INTENDED FOR GENERAL DISTRIBUTION
Never say, I know! Page 8 Eliyahu M. Goldratt, 2006

Do you know Robert Frosts poem The road not taken?

Two roads diverged in a yellow wood,
And sorry I could not travel both
And be one traveler, long I stood
And looked down one as far as I could
To where it bent in the undergrowth.

First, lets look down the meticulous road as far as we can (it is
meticulous and also boring, so please dont fall asleep before we start to
travel the other, much more exciting, road).

Shops might be reluctant to carry non-brand goods but it is a fact that
many shops do carry a lot of products which are non-brand products.
So, it must be that the managers of our company know that it is
possible to sell to shops outside their country but they don't really try to
do it because they are convinced that it might lead to losses rather than
profits. Let's carefully examine this conviction because the alternative
is simply to give up on the idea of selling directly to shops.

First, let's get rid of the more banal way. If the shops are reluctant to
buy non-brand merchandise because it is too risky, the company should
reduce the shops' risk by offering the goods on consignment. This is a
lousy suggestion. Offering the goods on consignment is very risky for
the supplier, since there is a high chance that most of the merchandise
will be shipped back at the end of the season. And it is bad for the shop
because, consignment or not, blocking a shelf with merchandise that
doesn't move well is reducing the throughput of the shop.

The shops that do carry non-brand merchandise are aware of the risk of
holding too much slow-moving merchandise. They reduce the risk
through lowering the price of those goods to the end consumer; they
increase the chance of sales of the non-brand products by fixing the end
price of non-brand products to be substantially lower than the brands
prices. But, at the same time, the shops are also making sure that their
margins will be adequate, which means they are willing to pay much
lower prices to the non-brand providers. How low can our company go,
and still make nice profits?

Currently, as we said, most of the sales of the company are not to
shops but to the brand companies. The current sales price of a garment
to the brand companies is (lets say) P, on which the companys
margins are about 0.25xP. The markup of the brand is about 400%
which means that the shop is buying that garment at 4xP. The markup
of the shop is about 100%, which brings the price the end consumer
THIS IS AN INTERNAL GOLDRATT GROUP DOCUMENT. IT IS NOT INTENDED FOR GENERAL DISTRIBUTION
Never say, I know! Page 9 Eliyahu M. Goldratt, 2006

pays for that garment to about 8xP (the margin of the shops is about
50% of the selling price). As we discussed, when the company sells
directly to a shop which is outside their country, the garment is a non-
brand garment and therefore, to increase the chance of selling the
sales-price for the end consumer must be substantially lower than 8xP.
Half the price can be reasonably called: substantially lower. That
means that to allow the end selling price to be just 4xP the company
price to the shop will have to be just 2xP. Since the totally variable cost
of the company is 0.75xP it leaves a margin of 1.25xP, a margin which
is 5 times bigger than the current margin. This straightforward
calculation suggests that the company could make a fortune by selling
directly to the vast number of shops located outside their country.

Lets not be too hasty to jump to such a conclusion. Remember, the
managers of the company do have a lot of experience and intuition.
Therefore, we can assume that they are oblivious to the obvious only in
cases where we clearly pin-pointed an erroneous assumption that
blocked them from seeing the obvious. In our case, did we identify
such a blocking erroneous assumption? No! So it must be that
something is wrong; that Ive ignored something essential in doing the
above calculation.

We are so used to getting much more from existing resources that for
us it is natural to first consider just the Variable Cost (TVC) just the
margin. But that doesnt mean that we are blind to the fact there are
circumstances where Operating Expense (OE) must be increased and
therefore must also be considered when evaluating impact on net profit.
Is the move to sell to shops in other countries associated with an
increase in OE? Of course it is. Sales costs will go through the roof.
Will they go up to the extent that it will consume the huge difference in
margin? Unlikely.

Are we sure? The expected margin is based on the assumption that the
likely selling price to the shops will not be less than 2xP. Is this crucial
assumption solid?

As we said, when the company tries to sell in other countries it is a non-
brand company; it does not have any real competitive edge. Are there
other non-brand companies that sell in those countries? Many. Many
non-brands, none having a competitive edge, create the ideal conditions
for the buyers of the shops to try to squeeze down the purchase price.
And they are experts in squeezing. Under the reality of a price-war
market, can we safely assume that the selling price to the shops will be
2XP? Lets ask it in a different way. Are the existing non-brand
companies making a fortune? Not at all. Some make a living, some
THIS IS AN INTERNAL GOLDRATT GROUP DOCUMENT. IT IS NOT INTENDED FOR GENERAL DISTRIBUTION
Never say, I know! Page 10 Eliyahu M. Goldratt, 2006

struggle, some are going out of business, but rarely do we hear about
non-brand companies in apparel making a fortune unless they
succeeded to build something specific that gives them a competitive
edge in a niche market. A price of 2xP seems a little too optimistic.

What is behind the bent in the undergrowth? The tendency is to still
do some checks. To check the price the shops are paying for non-brand
garments, to check what will be the cost of putting a sales force, etc. It
will take time, efforts and some investments but there is a chance that
as a result of such extensive checks the conclusion will be that it is a
(somewhat) viable road.

Now lets go the other way, the aggressive one.

As we said, the main reason that shops are reluctant to buy from a non-
brand company is that the risk that the merchandise will not sell well
enough is much higher than for brand merchandise. Lets not try, as
before, to reduce the gap between the respective risks. Instead lets
shoot for a much more ambitious target, a target that seems impossible
lets try to reverse the gap. In other words, lets try to find a way to
make the risk of the shop buying from our company to be much smaller
than the risk that shops take when they buy brand garments.
Aggressive? You bet. If we succeed then there is no doubt that we
have found an excellent solution. Choosing this way, lets concentrate
on the only remaining question. The big, very big, question: Is it
possible?

To answer this question we'll have to first assess the risk the shops are
taking when they carry a brand garment.

Is there any risk for the shop when it buys a brand garment? If we
think in terms of exploit the shelf space, there is definitely a risk. We
do know that even when we deal with brand garments there are a lot of
relatively slow movers, some are so slow that a shop has to carry them
for a few months and even then, for a major portion - about 30% of all
merchandise - the shop is able to sell (at a loss) only through the end of
season clearance sales.

Knowing the TOC solution for distribution the company can offer a much
better deal. Using replenishment, return for full refund and the
mechanism of return this, take that instead we can reduce the shops
risk to a bare minimum, while improving substantially the exploitation
of shelf space. This will ensure that (when the sale to the shop is done
in the SFS way) most shops, which are not dedicated to specific brands,
will accept the companys offer. Most probably, they will start with a
THIS IS AN INTERNAL GOLDRATT GROUP DOCUMENT. IT IS NOT INTENDED FOR GENERAL DISTRIBUTION
Never say, I know! Page 11 Eliyahu M. Goldratt, 2006

test collection but within weeks that test will expand creating a happy
(and loyal) outlet.

For those of you that are still not convinced lets do some calculations.
We know that the amount of shortages the shop has in the middle of
the season is about 30% (to prove it to the shop it is enough to take
the list of SKUs the shop is supposed to hold and compare it to the
actual inventory the shop has). When we use the TOC replenishment,
shortages drop to almost zero and as a result sales increase by over
30%. Yes, eliminating 30% shortages increases sales by over 30%
because the missing SKUs are the high runners (otherwise they would
not be missing).

On top of the sales increase due to elimination of shortages, sales
increase even more because of another reason. Conventionally, as we
said, the shop carries a lot of slow movers. When the shop realizes
what SKUs are the slow movers it will do whatever is necessary to sell
them. That means that there is great pressure to give the slow movers
more and better shelf space and more attention by the sales force.
Since shelf space is the constraint this is done at the expense of the
better moving SKUs. Using the TOC replenishment mode the amount
of slow-moving inventory is dramatically reduced. Less slow moving
inventory, less pressure to waste the constraint and therefore more
sales. Add to it the mechanism to replace slow SKUs with faster
moving SKUs and you get an effect on sales which is probably larger
than the impact of eliminating shortages.

Lets conservatively estimate the total increase in sales to be just 50%,
what is the impact on the shops profitability? Even though most shops
mark up the price by 100% the vast majority of the shops do not make
more than 5% net profit on sales. For such shops, a 50% increase in
sales means that their profit will increase by at least a factor of five.
With the understanding of such an impact do you doubt that: most
shops, which are not dedicated to specific brands, will accept the
companys offer.? Most probably, they will start with a test collection
but within weeks that test will expand creating a happy (and loyal)
outlet.

What about our companys investment? With such a mafia offer, the
company should choose to concentrate its sales efforts on a densely
populated area that can be served by one regional warehouse. The
amount of sales (potential sales are larger than several times what their
capacity can supply) and profits will easily dwarf the relatively small
investment in inventory.

THIS IS AN INTERNAL GOLDRATT GROUP DOCUMENT. IT IS NOT INTENDED FOR GENERAL DISTRIBUTION
Never say, I know! Page 12 Eliyahu M. Goldratt, 2006

We are not the first one to look for a solution to selling non-brand
directly to shops. Many, who are not less bright or less meticulous than
us, have tried to solve it. How come that we have succeeded where
they all failed? They tried to find the way to reduce the gap. We have
approached it differently. We first increased the challenge to the level
of impossibility instead of trying to reduce the gap, we dared to think
about reversing the gap.

Two roads diverged in a wood, and I--
I took the one less traveled by,
And that has made all the difference.