Spruce up that supply chain

Tech Talk

Every few years, management gurus come up with what may seem like fads that corporates latch on to.

In just the last two decades or so, we've seen several of these: from the more horizontal "just in time" manufacturing, business process reengineering and focus on core competencies to slightly more vertical concepts like customer relationship management, sales force automation and enterprise application integration. The scope of each of these ideas varies and has met with different levels of success.

Business process outsourcing (BPO) is the latest of these trends that we are living through. The idea is here to stay, never mind what John Kerry professes. With over 40 per cent savings and process efficiencies to be built upon, most, if not all, of the top global 10,000 companies have an offshore BPO strategy centred on India, the Philippines, China, Australia/New Zealand, Sri Lanka or even Trinidad & Tobago and South Africa.

But, very often, in the mad rush to please investors quarter after quarter, the same Global 1000 company managements, bitten badly by herd mentality, gloss over prospects for boosting revenues or contracting costs staring in their face. Such opportunities, admittedly, do not stare at you outside your kitchen window, but these are not rocket science either. The right amount of domain expertise, which companies that are leaders in their industry ought to have, mixed with clever innovation can deliver wonders.

Lead time optimisation, LTO in short, is one such very compelling concept, that U B Pravin and B G Srinivas, respectively, the heads of the retail vertical and enterprise solutions at Infosys Technologies, ran me through at a recent meeting.

Conceived primarily for the retail industry, LTO aims at making the supply chain more flexible for fashion houses, shoe manufacturers and other retail houses.

Squeezing efficiencies out of a supply chain is easier said than done, especially in the retail industry. The industry has done its share of spreading orders over multiple suppliers to optimally use their capacities; sourced orders from remote, cheaper locations (Nike in Thailand, remember?); and cut logistic costs by using a mix of ship, ground and air freight.

The problem of predicting demand precisely, however, has been a stubborn bugbear. Typically, the time from the design of a new line of, say, clothes to it actually reaching stores is up to a year. Supply planners constantly face the problem of being able to forecast how the new product will take off from the shelf. Sometime a new Gucci spring line may take off like hot idlis at an Udupi restaurant, but an equal "� or more "� number of times, they remain on the shelves.

The result? Companies often have to 'mark-down' "� discount sales in India and other parts of Asia "� their wares when they overestimate demand or have to shrug their shoulders with 'stock-outs' when they were overly conservative with their projections.

Infosys' Pravin, who is responsible for nearly one-eighth of the revenues at the software factory, told me mark-downs could account for as much as 30 per cent of the garment industry's sales and stock-outs another 10-15 per cent. Whew!

LTO aims at making the demand prediction software engine more accurate. It does this by analysing a sizeable number of variables like consumer preference, historical demand patterns and initial sales of a line (say, in the first week or two) region-wise, among others, to predict how a particular product will do. Tied in with the existing logistics management systems and other best practices in the industry (multiple vendor sourcing, tight inventory management), LTO can deliver phenomenal savings.

Infosys, which has partnered with SupplyChainge, a U.S. start up that has authored a better-precision predictive engine and decision support tool, is reporting good numbers. In a pilot project it has done for an American fashion retailer Pravin will not name, the lead time on a sourcing decision for a new line was reduced to three months from the traditional five months with 30 per cent better forecasting accuracy.

For a $1 billion apparel firm that loses annually up to, say, $100 million or $200 million on account of mark-downs and stock-outs, that is a fantastic addition to the bottomline.

"Typically, a company will be able to order a minimum quantity to reach its stores and depending on the performance, it could make its supply decisions," says Srinivas. "Fundamentally, [LTO engines] defer a buying decision. If you can delay by even a couple of weeks, your efficiency increases." Other names like i2 and Managetix also have LTO offerings.

LTO engines, which all of us will hear about more in future as more and more software companies woo their retail clients with their wares, work best with industries marked with two characteristics: high unpredictability of demand and long lead times in the supply chain.

Obviously, then, the highly fickle apparel and footwear industries are prime candidates. Other industries "� mid-range and high-end wines or jewellery, to name two "� could also perhaps have use for this concept.

(Josey Puliyenthuruthel works at content company perZuade. His views are personal and may not be endorsed by his employer, the company's investors, customers or vendors. Comments may be sent to josey@perzuade.com)