A CEO's experience is important - the Difficulty at The Gap in 2007

I love BabyGap. I think the designers do an incredible job. The product looks fresh and cute, and it is priced reasonably. I read a very interesting article in February's 26th Business Week, Paul Pressler's Fall From The Gap. My first reaction was here is someone who didn't read Kathleen's book. Reading through the article (and if you can believe everything in it), one can see he clearly didn't understand the fashion business. It would be easy to put all of the blame on him because he was the new guy. The truth is that GAP has corporate culture issues not unlike any large company.

I met an assistant designer for GAP years ago. The impression she left was not very favorable. At least I decided I would not work there, if given the choice. It was a pretty cut throat environment with people climbing the corporate ladder rather quickly because of constant turn over. She left the impression that she could be become one of the head designers within a few years. Success like that does not happen without a lot of back-stabbing.

So I don't blame him entirely for his failure to turn things around at GAP. It would not have been an easy task to walk into a difficult corporate culture, with little fashion experience. Some of his decisions illustrated in the article were indeed poor, others just missed the mark. I think design entrepeneurs can learn from him. Here are my thoughts on some of his decisions:

1. Combining fabric purchases. He made the mistake of combining denim sales for all four divisions of GAP, including Old Navy, Banana Republic, and Forthe and Towne. The result was that all divisions ended up with exactly the same denim to use in all of their denim styles. There is no point in having four divisions if there is not some kind of difference in the clothing. Why buy denim jeans at Banana Republic when you can get the same thing at Old Navy for less money?

The idea is not without merit. I have had corporate execs suggest the same in the past. The goal is to negotiate a lower price for the raw materials by buying in bulk. A small company can combine fabric purchases to also negotiate a lower price. The point is that you shouldn't buy only one fabric, or one style of fabric. Large fabric mills will push you in that direction because it is less expensive for them to run 10,000 yards of the same denim fabric. But what are you going to do with 10,000 yards of denim?

A small business is more likely going to need, say, 300 yards of denim, which will cost more than what GAP will pay. But you can still negotiate a lower price. Sometimes, your order can be combined with an order from another company. Both companies could benefit with a lower price. Perhaps, you could up your order to 500 yards and you can use the extra fabric the next season or in another style. If the fabric company has more than one type of fabric, perhaps you could combine your fabric needs with one company. A long term relationship between supplier and buyer can lead to lower prices.

A small business should never buy ALL of their fabric from one source. You never know what could happen, so make sure to have back-up choices.

2. Outsourcing development. Pressler required his designers to create their patterns in the states and then have the samples made in Asia. Sample making in Asia is cheaper. The problem is time. It takes time to mail samples and patterns back and forth. Even with internet technology, there are many barriers to the ease of communication. Language differences is the biggest. This kind of product development can take as long as 3-4 months for final product approval, especially on a new style. Pressler would not give approval for expedited shipping.

You can outsource product development. It should be something to consider, especially for childrenswear designers. A lot of children's clothing is made overseas - just shop the competition and you will see very few USA labels. Because competition is so stiff, you can save a lot in labor by moving overseas.

If you decide to take this road, give yourself plenty of time. Time to find a reputable manufacturer. Time to find a manufacturer that will run smaller lots (small is a relative term, but expect 500-1000 pc runs to be termed "small"). Time to teach the manufacturer and their technical people your product and quality standards. You can do a lot of things long distance. One Chinese manufacturer preferred that designers actually come and spend a few weeks in China working directly with a patternmaker to develop samples. Not only was it faster, but it ensured the designer got what they wanted.

3. Market Research. Pressler tried to use traditional market research tactics to predict the next big trend. The problem is that customers look to their favorite brands to lead on the next trend. This means that designers predict or envision what should happen next. Fashion companies rely on their designers for this.

Market research is important for building a better brand and better product. Designers do need to listen to their customers. I have seen a few online fashion companies add product reviews and customer comments on their sites. This type of system can help a company improve problems faster. Inspect returns for product defects. Learning from the past can help you move forward.


What will happen at the GAP next? They are now on the hunt for a CEO with apparel experience. I don't see the death of GAP, but I do see some serious growing problems.

Another Care/Content Label Example

It's time to look at another care/content label. These labels were found in a 12M knit top that is part of a two-piece set.

The first thing that caught my eye was the placement of the country of origin. Country of origin information must be placed so that it can be seen without flipping the label over. Nothing can cover the country of origin information, like a brand label. It should be placed in an easily accessible area, like the back neck or waistband.

Another interesting observation is the style number and lot number information included on the front of the label. The lot number probably refers to the production lot and corresponds to a date of manufacture (lot numbers are unique to each company, so it could mean almost anything). Generally, this information is found elsewhere, like the reverse of the tag or even on a separate tag stuck into a side seam. It is certainly not anything that a customer is going to care about. I like the inclusion of this info for easy back-tracking, but it should be placed elsewhere.

This tag contains both an RN and CA numbers. The CA number is the Canadian equivalent to the the US RN number. This product was most likely sold in both Canada and the US, thus the need for two numbers.

The reverse side of the label contains all relevant information. My only nit-pick is the overall formatting. I generally prefer the content information first, followed by the care info. This label is just the opposite. Further, the content information is a different font, font size, and right-aligned compared to the care instructions. It makes the whole reverse side appear sloppy.

My final nit-pick is that the label should say "2 piece set" somewhere. Inevitably, the two pieces will be separated. This helps the shipping and store employees on the retail end realize there should be two pieces on the hanger.

Grading Stack Points

I am going to try and get back to my grading explanations today. My head is a bit fuzzy after a long week and an especially late night. To review, just click on the grading category link to the right. Click on the picture for a bigger view.

The diagram on the left shows three possible grading examples. This is a basic front bodice pattern piece that is folded in half. Each example is called a nest meaning each size is stacked on top of each other from smallest to largest. The stars represent the point of origin for each grade. The numbers at each point correspond to a grade rule. I will briefly explain each of these items below.

By laying your graded patterns on top of each other in a nest, you can easily see if your pattern is grading proportionally. The stars not only show the point of origin, but the stack point. You can stack your patterns in three general areas and have the grade look different each time. In computerized grading, you can move your stack point anywhere you choose and the grading software will update your grades automatically. For consistency's sake, it is important to determine before you start grading where your stack point/grading point of origin will be.

Each stack point will cause your grading to look different. Most of the grading world uses Example 1. The point of origin is placed in the center of the pattern pieces and growth occurs in all directions. In Example 2, the stack point is placed at the center front waistline and growth occurs to the side and upwards to the neck. Example 3 is the same result as 2, but placed at the center front neck.

I prefer Example 3 and all of my patterns are generally graded this way. It makes the most sense to me and the grade rules are simpler. Most of the proportional changes between sizes occur this way in nature. A person's neck doesn't actually move upward in each size, but rather their waist moves down. Perhaps this is just how my mind works. In any event, you can choose which ever point makes the most sense to you.

I have been using the term grade rules rather loosely in my previous grading articles. Grade rules refer to measurement charts broken down into grade steps. Grade rules can also refer to the actual change that occurs at a point. In the drawings above, I have numbers assigned to each point. Those numbers can refer to a grade rule. For example:

Rule X,Y changes

1. 0, 0
2. 1/8, 1/8

You can create a chart like the one above, if you prefer. This method is used in computerized grading and each grade rule is placed in a grade library (specific to a size range). Once a style is ready to grade, you simply apply a rule to each point and the pattern pieces will stay graded regardless of pattern changes. It is less helpful when hand grading and may make things more complicated. To be honest, I rarely use grade libraries when computer grading. They can speed things up if you have consistently similar styles. However, things always need tweaking, so I prefer to manually manipulate the grade at each point. Most grading software will allow you to assign a grade rule from a library and still manually edit the point. The computer will begin a grade command by starting with rule #1 and working around through each point. Rule 1 is almost always set as (0,0) and it should always be your stack point.

When hand grading, you should start your grading at your stack point. Once that is set, you then move to the next point, say the neck-shoulder point and mark all of the changes there. Work counter-clockwise (or vice versa, if you choose), around to each point. Setting up a consistent method will help you keep things straight.

Ok. Enough explanation. Decide on your stack point and get ready to grade.