Pull Based Replenishment

Replenishment has undergone a swift change in the past 20 years or so with the focus now shifting to replenishing only what is actually consumed or the off-take from POS.Supply chain strategies are being re-engineered and business processes re-molded to have strategic alignment.Not surprisingly, P&G is the leader of such a movement in Indian FMCG sector with more than 90% of the total sales committed through such a system.Though others have also followed the suit but with less success.

FMCG is one of those sectors which is highly consumer centric and the nature of demand is independent.The demand is in fact dependent on a lot of factors other than what the replenishment managers or the business forecaster can predict.Typically the replenishment in a push based supply chain is on the basis of past sales and sales target along with the inputs of supply constraints.But all these do not take into account subjective factors which marketers highly recommend such as taking consumer psychology and shifts in preferences over a period of time.Also the competitive factors,product life cycle and introduction of variants and new SKUs influence the actual demand.The introduction of a competitive product may skew the entire rationale behind the forecast.The result of such a replenishment system is excessive inventory of slow moving goods and stock-outs for the fast moving.The excessive inventory forces the firm to provide discounts and higher margins to distributors and retailers in order to push sales.Such discounting leads to loss in profitability.The stock-outs which obviously count as monetary loss but more than that it can be detrimental to the brand image and the consumer loyalty may shift towards the more easily available substitute.In order to compensate for the lost sales in the previous period the forecaster may increase the forecast for the coming period but in actual the demand is substituted by another product thus this time leading to excessive inventory.This can be a cascading effect with the forecaster actually having no sense of the actual demand.Hence it is very difficult to quantify the actual demand that a product has.Few companies have tried to improve the forecast by involving more and more factors while forecasting.Most of these efforts leads to complex mathematical models which over a period of time becomes unmanageable.

Few other firms have tried to move their replenishment system closer to the customer and have tried to capture the actual demand and consider that as a baseline for future sales rather than sales target.The first step towards this approach is to capture the POS data.The Indian Supply Chain presents its complexity at this very first step.With the Indian Distribution System of small Mom & Pop stores,its a challenge to get the data from each of these store,for many of these have turnover less than the money required for setting up such an infrastructure.Hence most of the firms capture data at AW level.For a typical FMCG number of distributors are in excess of 5000 but the retailers may run into tens of million.  

With the POS data available and future growth pattern identified the forecast is statistically more correct.The statistical forecast is then improved by taking into consideration the promotions and brand marketing initiatives to be launched in future.Generally the short term forecast is done for a period of three months which is a rolling forecast and is updated as the actual sales happen.The rolling forecast provides some flexibility in terms of correcting for the actual trends emerging from POS data.This is the first step towards the pull based supply chain.The backbone of such a system is a robust,dedicated and real time IT infrastructure so that the information reaches the required person at right time.With the value added forecast available it becomes necessary to plan the production accordingly.Most of the companies plants that are outsourced are located across the country hence it becomes imperative to decide which SKU needs to be produced at which plant in what quantity to be supplied to which market.Here comes the role of the planning department.After this planning even the production schedule is mapped as the demand moves and fulfills just the quantity needed while taking into account safety stocks and other constraints.Finally the dispatch plan is prepared according to the sales trend.Thus we see everything gets interconnected and the product is pulled through and it travels through the entire value chain as per the trend and demand of the market.Demand induces forecast which gets transformed into the production plan in accordance to the market dynamics and not sales target.The Production plan decides the dispatch plan and finally the product is delivered.

In my view above product flow presents two important facts to be considered.One is that the setup requires integrated or collaborative supply chain approach and the other is flexible factory operations for successful implementation of this pull based approach.Until the supply chain members do not understand the correlation and interdependence of operations it will be difficult to operate such a supply chain smoothly.Identification of bottlenecks and improvement of the same should be a continuous process as the business expands and few nodes become critical.The outsourced plants presents even a higher challenge to match such a replenishment system because of the contract clause and agreed terms and conditions.Flexibility both in terms of product variety and batch size is important with minimum changeover time and cost.The quantum of success will depend on degree of compliance of these two factors.

It will not be wrong on my part to presume that by crashing the planning horizon within which we can extract the demand information, prepare a forecast, develop the production and dispatch plan and deliver the goods will move us closer to the actual demand fulfillment.This needs to be corroborated further but reducing the time uncertainty will certainly improve the results.