Vendor managed inventory (VMI) is sweeping through many areas of retail as the next step in supply chain management. Retailers, particularly those in commodity markets, see it as an essential technique to reduce inventory - and apparel sellers are no exception. Although VMI is usually practiced between a manufacturer or middleman and a retailer, it can also automate just-in-time delivery of supplies to the manufacturer. This article by Prabir Jana, Professor Prabir Banerjee and Dr Alistair Knox explores the potential of VMI in apparel manufacturing.

In principle VMI increases sales by avoiding out-of-stocks on in-demand products, and reduces losses from overstocks of products that no longer sell. Although VMI is usually practiced between a manufacturer or middleman and a retailer, it can also automate just-in-time delivery of supplies and parts to the manufacturer by its suppliers [1].

Although VMI is practiced in the electronics industry between final assemblers and part suppliers [2], there is no evidence of VMI between apparel manufacturers and raw material suppliers. This article explores the potential of VMI in the apparel manufacturing environment with supporting case studies.

What is VMI?
Vendor managed inventory is essentially a distribution channel operating system whereby the inventory at the distributor/retailer (dist/ret) end is monitored and managed by the manufacturer/vendor (mfg/vend) [3]. It includes several tactical activities including: determining appropriate order quantities, managing proper product mixes, and configuring appropriate stock safety levels.

The rationale is that by pushing the decision-making responsibility further up the supply chain, the manufacturer/vendor will be in a better position to support the objectives of the entire integrated supply chain resulting in a sustainable competitive advantage.

Through access to the retailers' sales data manufacturers can produce in response to market demand, reacting faster to changing needs. Interestingly, the technical support needed for VMI - once EDI is in place - is actually fairly straightforward. It is the relationship that requires effort [4].

How is VMI different from traditional inventory management?
VMI is typically the opposite of the inventory management approach taken by most organisations today.

Currently, orders are pulled through the supply chain by each partner as inventory levels reach "replenishment/re-order" points. This historic approach only serves to spur channel partners into optimise their individual link in the supply chain at the expense of sub-optimising the overall supply chain.

VMI, on the other hand, works in reverse to link partners together and to grant authority to the partner who is in the best position to make inventory replenishment decisions. This entity is usually the mfg/vend partner given its upstream position in the channel. The overall goal must be to support total value chain cost minimisation by pushing decision making on replenishment activities furthest up the supply chain.

Why VMI?

  • To reduce/share cost of inventory
  • To reduce/share space requirement for inventory
  • To cut down response (procurement) time
  • To reduce/share uncertainty of type of merchandise requirement
  • To reduce/share uncertainty of quantity of merchandise requirement

Typical benefits to manufacturers

  • Lower inventory investment (raw and finished)
  • Better scheduling and planning
  • Better market information
  • Closer customer ties and preferred status

Typical benefits to distributors/retailers

  • Fewer stock-outs with higher turnover
  • Better market information
  • More optimal product mixes
  • Less inventory in channel (transfer costs)
  • Lower administrative replenishment cost

Source: [1]

Successful VMI implementations

  • At K-Mart, customer service measures have gone from the high 80s to the high 90s. Inventory turns on seasonal items have gone from 3 to between 10 and 11, and for non-seasonal items from 12-15 to 17-20.
  • Fred Meyer, the 131-unit chain of supercentres in the Pacific Northwest, reduced inventories 30-40 per cent, while sales rose and service levels increased to 98 per cent. This was due to a VMI programme implemented with two key food vendors.
  • Wal-Mart and P&G have had a VMI program together for over ten years to manage the inventory and production of disposable diapers, with great success: turns doubled, Wal-Mart's operating costs fell, and P&G's market share grew (because Wal-Mart gave it preferred shelf space).

VMI applications in apparel manufacturing
VMI can be looked on as a step beyond just-in-time (JIT) in the manufacturing scenario.

In JIT, the vendor is supplying inventory to the manufacturer 'just-in-time,' whereas in VMI the vendor is maintaining (it has already supplied before time) inventory at the manufacturer's warehouse.

In practice, JIT pushes the effects of unpredictability upstream in the supply chain [5]. While the manufacturer is required to predict (forecast!) retail demand, the consumable vendor further up the supply chain can simply calculate the manufacturers' requirement. This is the key difference.

The quality and quantity of merchandise in VMI are two very important (and often disguised) parameters for its application in manufacturing. Unlike the retail scenario where inventory has an individual identity and is tracked by SKU, inventory at manufacturing level has two identities: quality and quantity. Predicting inventory at the manufacturing level involves getting both quality and quantity right.

Some of the areas of potential VMI application in apparel manufacturing are consumables procurement (i.e. the raw materials and accessories) shown below:

  • Fabrics
  • Sewing threads
  • Needles
  • Zippers
  • Sewing machine spare parts

The following table summarises the potential of each of these areas against different VMI parameters.

click to enlarge
click the table to enlarge

For example, practising VMI between the needle supplier and apparel manufacturer will result in negligible cost and space savings for the manufacturer. Also, the manufacturer felt that the current response time (procurement time) by the needle vendor was satisfactory and further improvement would not result in any substantial benefit.

VMI is also about the accurate prediction by the vendor of merchandise quality and quantity. While merchandise quality (type of needle) can easily be predicted (certain), the quantity (of needles) is uncertain and needs expert calculation by the needle vendor.

In the retail environment one of the keys to VMI implementation is the successful prediction of the right quantity and quality of merchandise. A sewing thread supplier can decide on the type of thread and calculate fairly accurately the quantity required based on the fabric swatch and apparel sample.

The question, however, is whether VMI application will result in any further advantages in environments where the prediction of quality and quantity of merchandise is relatively easy (just calculation, no uncertainity!).

Taking VMI in the retail environment as a benchmark, similarity and disparity analyses for the five consumable supply environments listed earlier were carried out. The comparative similarity or disparity suggests the respective ease or difficulty of implementation.

click to enlarge
click the table to enlarge

VMI case studies in Indian apparel manufacturing

Case one:

One apparel manufacturer exporter from Bangalore wanted to experiment with the VMI concept primarily for controlling sewing thread inventories. Sewing thread is a very important item in apparel manufacturing. The demand pattern (based on the production schedule) was shared with the thread manufacturer and the mandate was to supply the right quality and quantity of thread at the right time.

The thread supplier, in consultation with the apparel manufacturer, calculated the actual quantity of thread. Before implementation of the VMI concept the manufacturer alone would calculate and order the total quantity of thread, with any surplus or shortage being the manufacturer's problem.

Under the new concept the onus of any surplus or shortage was shifted totally to the thread supplier. Somehow the thread quantity calculation was never found to be accurate (despite the best efforts of both parties), so the thread supplier had to bear the cost of surplus thread.

Barring some standard colour like white or black, thread is a perishable item so the thread manufacturer withdrew from the arrangement. The apparel manufacturer went on to have a risk sharing agreement with another thread supplier and continued to work on the concept. Under the new risk sharing arrangement the cost of dispute/surplus quantity of thread is being shared by both the thread supplier and apparel manufacturer.

Case two:

An apparel manufacturer in NCR (the Northern Capital Region) has a just-in-time delivery arrangement with a thread supplier. Under this arrangement the manufacturer blocks certain quantities of thread with the thread supplier, who then maintains the thread inventory and delivers products just-in-time when required by the manufacturer.

Blocking thread quantities is only valid for a certain period of time, after which the thread supplier is free to sell the thread to other manufacturers. This arrangement is only possible because the geographical proximity of both companies guarantees delivery within 8 hours.

Under this improvised arrangement:

  • The thread supplier avoids complicated reverse entry of goods if there is a surplus
  • The manufacturer avoids being left with surplus stock, while guaranteeing against out of stock
  • The responsibility for calculating the correct quantity still remains with the manufacturer.

The analysis
Other products like zippers and needles were also explored but manufacturers felt these were too insignificant to experiment with. Needle and zipper suppliers, however, showed an interest in investigating further.

One manufacturer showed an interest in setting up a VMI arrangement with a yarn and fabric supplier and the possibility is currently under discussion. CAD/CAM solution provider Gerber Technology launched a VMI programme in 1999 [6]. Under this pilot programme, Gerber's service organisation partnered several customers with large cutter installations to maintain a new parts inventory at the customers' site, resulting in maximised customer "uptime" and reduced service related costs for Gerber.

While exploring similar arrangements with sewing machine spare parts suppliers in India it was found that similar practices are being exercised with a few large apparel manufacturers for selected high priced machines.

Two reasons are preventing VMI from becoming common practice. Firstly there is an absence of large apparel manufacturers, the driver of VMI implementation. And secondly, there is an absence of major machine manufacturers.

According to Professor Sunil Chopra, VMI should only be implemented in cases where the manufacturer/vendor can forecast demand more accurately than the distributor/retailer.

Accessory and raw material suppliers are better able to forecast (sometimes just calculate) demand from apparel manufacturers, whereas it can be difficult for manufacturers to forecast demand from retailers.

Fabric (basic raw material) and sewing threads hold specific challenges due to the unpredictability of demand, whereas needle, zipper, machine spare parts etc have fairly stereotyped demand patterns and can be easily implemented.

VMI presupposes EDI between the trading partners, since an absence of the EDI infrastructure will make the "time gain" factor difficult to appreciate (and convert to a cost advantage) by partners.

However, the emphasis is on the relationship, and the software merely automates the demand analysis. The sales tax and other procedural complexities may need to be simplified if there is to be a smooth flow of material and information between partners.

This article is based on annual research carried out at the GMT Department, NIFT, sponsored by Creative Garments and with technical assistance from Creatnet Services, India. This is part of doctoral research by Prabir Jana at Nottingham Trent University, UK.


  1. Fernando del Cid, Roger Gordon, Brian Kearns, Paul Lennick, and Andreas Sattleberger under the supervision of Sunil Chopra, Professor of Operations Management, J. L. Kellogg Graduate School of Management, Northwestern University.
  2. Duncan, R. Logistics: Trading Places. Engineering, Nov'96
  3. "State of a New Art." Manufacturing Systems (Master the Supply-Change Challenge Supplement), pp.: 2-10, August 1995.
  4. G. BERTON LATAMORE, Customers, Suppliers Drawing Closer through VMI Relationships, APICS - The Performance Advantage, July 1999, Volume 12, No. 7
  5. Mallman, D.L., "The strategic role of suppliers in manufacturing strategies", proceeding of the 2nd international symposium on logistics, Nottingham, 1995, pp.3-8.
  6. News Release, JIAM Show 1999, Gerber Technology