The operating environment for businesses inthe Apparel & Footwear Industries has changed in recent years. Global trading andcompetition, low margins and more orders in small quantities are characteristic of today'sIndustries. All of these have created pressure on domestic businesses to grow strong ordie. The application of IT in the form of fully integrated Enterprise Resource Planning(ERP) systems has been a key part of the success of the thrivers. After all, how can youcontrol costs and maximize sales unless the activities of all resources are integrated?Business units must keep in touch with each other and the marketplace. These systems havenow become the norm. The question to ask is: are they all the same?

Any salesperson will concentrate on themerits of his/her own product, and will be unlikely to spend much time on its weaknesses.It is crucial for those involved in system evaluation to look hard at system fundamentals.How will it work for our business, with real people, real problems and using real data? Ifa system is developed, either generic or for a particular industry, it will not suit theunique requirements of apparel and footwear businesses. It is, inevitable that softwarepromotions tend to concentrate on high-level, strategic issues. Sometimes, the firstopportunity that users get to try the system is after it has been bought.

One of the justifications most often quotedfor the implementation of ERP systems is that there will be '....significant savings inadministration....'. The speaker intends the prospective purchaser to believe that forless input there can be more output using the system. Indeed, this can be true. However,it does not come about by accident. It will result only if the system is designed tooperate in a way which suits the industry in question. It must make sure that system usersact more intuitively and confidently. It should speed up work and leave more time forbusiness management.

Data entry complexity

'Businesses run on the details' : This is atrue statement. Irksome though it may be users must enter information into the IT systemin detail. Having done so, we must use, maintain and archive or delete all of thosedetails. This data can be summarized into management information. However, without thedetail, there are no reports and no automatic processing. This article will examine oneaspect of the administration of an integrated system in order to prompt discussion of suchoverlooked areas.

One of the most basic of all administrationactivities is data creation for styles, Bills of Materials (BOMs) and routes. Let us beginby determining what sort of numbers we might be dealing with. i.e., first, the number ofStock Keeping Units (SKUs) required. One example could be for a pair of women's trousers:

  • the trousers have a style or stroke number, T5678
  • they come in a range of four colors
  • they are made in ten sizes, with four fittings

For this one style, the number of SKU's is160 (1 style x 4 colors x 10 sizes x 4 fittings). This number, however, reflects only thedata required to record the finished product options on one style. The products have to bemade and raw materials have to be purchased. For the system to handle these processesrequires that data be entered for the BOMs and the routings. The administrator must thenset up system links, or associations, between the materials, the routes, and the relevantSKU's.

If we assume that the above product istypical of what goes on in our hypothetical business; that the typical product is madefrom ten purchased parts, via a sequence of 20 operations; and that we have 200 styleslive at any one time, the picture changes dramatically.

In this fictional, but typical scenario,users are faced with the entry and maintenance of 25,600 BOMs. Further, they must create256,000 relationships between these BOMs and the raw materials, plus 512,000 links betweenBOMs and routings.


Fig.1 An example of Bill of Materialsand Routing Associations

With some ERP systems, this data would haveto be manually created and linked to the relevant SKU's. An additional problem is wherethe SKU and the associated data have to be recorded against each finished goods stockroom.In this case, we may have five stockrooms, and consequently five times the data. Also, theaccuracy of the base data and all subsequent data manipulations depends entirely upon theaccuracy with which these processes are carried out. How are users going to ensure thatthey achieve the level of data accuracy that is required for the system to functioneffectively? If they are unable do so, the system may create more problems than it solves.

Most systems allow copying of data whichwould save some time. However, even with a copy facility, the number of screens which haveto be opened and closed in order to move the data around means that manual entry of thisamount of data would be both time-consuming and prone to user error. In addition, amistake made on version one would be copied across. In the same way, what happens if apiece of data common to many SKU's changes? For example, material usages can be revised oroperation timings could change.

This is a serious problem and directlyeffects the efficiency and effectiveness of a business. Today, systems are being sold andin use in which each of the entries must be separately maintained. If that is the way asystem works, and a business is in an industry where the products have characteristicslike style, color, size and fit, then that business may need to recruit some extra staff.

This example illustrates what the systemadministrator must do to facilitate the most detailed planning and costing scenarios. Insuch systems, therefore, attaining this desired level of detail is effectively preventedbecause of the time and effort involved in system set-up and maintenance. To overcome thelimitations of their systems, companies may opt to apply their BOM and/or route at stylelevel. The advantage of this approach is that administration is manageable.

However, circumstances often arise when theability to apply different machines, labor costs, or material usages between sizes andcolors is necessary. It seems that systems providers commonly offer the choice of twoextremes - simple administration but limited detail, or all the detail but unmanageablesystem administration. Neither achieves the desired outcome. With a little lateralthinking, however, systems could provide the best of both worlds. For example, if BOM androutings were defined at style level with changes to operations and raw materials definedonly where they differ (color or size level) we could achieve the ideal of simpleadministration combined with all the desired detail.

It could be argued that the move from twodistinct seasons to an environment in which retailers are looking to constantly refreshtheir ranges may help deal with this problem. New style creation will become more of acontinuous process and less of a mad rush every six months. However, this smoothing of theadministrative burden has been far outweighed by the dramatic increase in the number oforders, often for smaller quantities. This increase in the number of products beingmanaged looks set to continue, making product data management more complex.

How might a system that was conceived todeal with the specific requirements of the Apparel & Footwear Industries deal withthese complexities? Possibilities exist for the automatic entry of some of the data. Forexample, some ERP systems can be interfaced to software that facilitates the electronicmanagement of the design and development areas. This, in turn enables the product data forthe base size to be automatically transferred into the ERP system. Even under thesecircumstances, however, there remains a great deal to do in order to fully load the datafor all SKU's and create the BOM and routing associations. What is also required is ameans of entering, viewing and reporting that works like the people in the Industry.

A typical order breakdown by quantity,color and size in the Apparel & Footwear Industries, will look like Figure 2. Figure 2shows a typical style/color/size matrix. It is a concept common to these Industries. Ifthis concept were applied to data entry, display and reporting, the possibilities forimproving the situation would be enormous. Effectively, this would create a grid withsizes and colors.


Fig.2 Style/color/size matrix

If a screen could be accessed that alloweddata entry, display and reporting via this cross-referenced system, there would be anumber of major advantages:

  • The screen would make immediate sense to the person making the entries. When finished, the result would be presented in the same way as the original information. This would lend itself to easy checking.
  • If all subsequent screens and reports were presented in this way, a common reference point would be available for both internal and external communications. The data would be available in the language of the industry.
  • The possibility would exist for a screen to be displayed quickly based on the known coordinates defined to the style. So, with ten colors and 12 sizes, the screen could display the 22 (10 + 12) attributes rather than accessing the 120 cells that individually comprise it. This approach would make the system extremely efficient and users would have faster operation in all aspects of system use.
  • Using such a method of entry and presentation would ensure that the sizes appear in the sequence preferred by the users and not in a sequence based on rigid system logic. For example, if a system requires that colors and sizes are entered by SKU, the information will typically be displayed in alphabetical order. For instance, if you handle small, medium, large and extra large, with system codes of S, M, L and XL. With most systems, the screen display and any reports will read L, M, S, X. This may seem like a small point, but it becomes important when reviewing stock to maximize sales by considering size and color ratios.

Also, what if a style is carried over to anew season but made in a new range of colors? You would then have stock of the oldseasons' colors and would need to display that stock on screen with the new, in order thatit is not forgotten. However, the new colors need to appear at the top of the matrix sincemost work is being done on these new colors. They will not do so if their position isdependent on system logic.

  • Application of such a concept would prevent invalid or erroneous entries. If a particular style was not to be made in certain colors or sizes, entry could be prevented by setting a warning flag at the relevant coordinate.
  • System operators could access data by working at style level. This would drastically reduce the number of codes which the administrator must know. The color codes would not need to be known. The information could be accessed at style level. The opportunity could then be given to drill down to the lower levels. The major advantage of this would be the need to access only one screen, making the system faster and easier to use. This would be particularly helpful for infrequent users who may not know their way around. It would be a simpler and much less frustrating approach to use. This problem also applies to partial matrix systems that still require color codes to be entered.

Figure 3 represents the sizes and colors ina shirt style S1234. At the point X, a relationship could be created within the systembetween the color and size characteristics. For example, it could be that the largest twosizes (42 and 44) are technically different to the smaller sizes. For this example, sizes42 and 44 each requires extra fabric, one more button and five Standard Minutes extramachine time on a different machine. If the standard BOM and route at the highest (style)level, in this case 'S1234', was applied, it would give the minimum possibleadministrative load. We could then specify the differences in fabric and buttonconsumption plus the additional Standard Minute Value of the operation and the change ofmachine at the points in the BOM and route where they occur. Administratively, this isvery straightforward. Each time data is entered for sizes 34, 36, 38 & 40, the systemdefaults to the standard BOM and route. For sizes 42 & 44, however, the system willapply the exceptions in the BOM and route which we specified. Purchasing, planning andcostings can now all be carried out at the finest level of detail.


Fig.3 Sizes and colors of shirt styleS1234

The simple example in Figure 4 proposes acombined BOM and Route. These are standard at operations 020, 030, 050 and 060. Thevariations occur at operations 010, 040 and 070.

This application of a 'default' BOM androute at the highest level, combined with the ability to invoke exceptions as requiredthrough the use of characteristic relationships, could equally resolve the requirements toprocess different colors in different ways. Indeed, any exceptional change to a BOM orroute could be handled without overloading administration.


Fig.4 Combined BOM and Route

In summary, those involved in theevaluation of ERP systems must appreciate the importance of system design. All systems arenot equal. This is particularly true when considering the requirements of businessesinvolved in the Apparel & Footwear Industries. These Industries have uniquerequirements and systems must be designed with their specific needs in mind. Thedifferences have implications for data entry, display, reporting, and processing all ofwhich could be made much simpler and faster using an industry-based approach.

Conclusion

ERP systems are intended to enablebusinesses to proactively use integrated information and resources to improve processesand optimize business performance. All business reports and forecasts depend upon theaccuracy and accessibility of data. The complexity and workload involved in setting up andmaintaining such data is usually both underestimated and undervalued. The methods ofentering and presenting data available in systems that were not designed for the uniqueneeds of the Apparel & Footwear Industries are unsuited for these Industries. Often,those responsible for system purchase are given no warning of the difficulties that mayarise. Neither in terms of accuracy, nor in terms of workload. Only once implementationhas begun, does the problem become apparent. Often it is then too late.


Fig.5 Comparison of administrationtimeline

What is required is a method of data entry,display and reporting which is quick and easy to use. It must offer the maximum amount ofinformation from the minimum number of screens. The data must reside in a singlerepository for system efficiency. The system should be completely integrated, facilitatingup-to-the-minute data availability. In this way, time spent on administration can bereduced and data accuracy increased and maintained at the required levels. To facilitatethe best possible management of the business, more time must be spent on using accuratedata to produce quality reports and forecasts. These issues are a major problem in systemsbeing sold and used today. They can only become more important as businesses grow and thetrend towards more products in more ranges continues.

Robert Jennings is product marketingmanager for JBA Style. He has worked for international companies such as Coats Viyella,Skopos Design, Celestion Textiles and Claremont Garments in a variety of roles includingproduction planning, purchasing, supply chain development, materials control, productsourcing and sales. He also has five years experience of retail store management and is aTextiles and management graduate of The University of Leeds.