Finding the single version of the truth when it comes to sales figures is a challenge for any organization. The marketing department calculates differently from finance or buyers, while salesmen have a spin entirely their own.
It’s the same when it comes to predicting demand.
The store manager has a pretty shrewd idea but adds on a wee bit just in case. The distribution centre aggregates the stores’ suggestions and rounds up just to be sure. And as for the buyer – well, the supplier has a fantastic deal for only a few more cases than the forecasts suggests.
And so it goes on along the supply chain. This tendency to add extra safety stocks combined with the failure to take account of what is already on outstanding order or in transit pushes up inventory levels from retail shelf all the way back to raw material suppliers.
Core consumer demand
Just as IT systems have finally brought those multiple sales figures under control, so the exponents of ‘flowcasting’ argue for a single demand forecast based on store-level needs and then accurately transmitting that forecast along the supply chain. The forecast takes store level demand on a day-by-day basis over a defined period and the system calculates how much must be shipped and when, based on this core consumer demand forecast. The flowcast model is then used to drive financial projections for sales and payments to guide cash flow planning as well as calculate warehouse requirements over time.
It’s an approach which demands collaboration and trust among trading partners as well as greater flexibilityin planning with retailers willing and able to predict demand 15 weeks or more ahead and share that information with suppliers. It is also an approach which, it is claimed, can add up to six per cent to the bottom line and cut inventory levels by up to three days.
‘Forecast accuracy is typically plus or minus 20 per cent,’ suggests Mikael Bisgaard-Bohr, retail industry director for NCR Teradata’s European operations.
‘Extrapolate that for all the forecasts as you go up the supply chain or to other parts of the business, and you can see the scale of the problem. Safety stocks can easily amount to 10 per cent of total inventory. Cut that even slightly and performance can be improved dramatically.’
Bisgaard-Bohr argues that most replenishment systems are based on a point in time. By adopting a more dynamic approach looking at the total demand pattern for a product throughout the supply chain over, say, 15 weeks you achieve a more accurate picture of real sales and product needs. Teradata calls this ‘timephased replenishment’ and has included the technique in its demand chain management application for some time, but the approach is close to the flowcasting model now being adopted in the US.
Evant – acquired by Manhattan Associates last year – has a similar model, this time called ‘multi-echelon replenishment’.
‘Evant always used this approach,’ says Tim Thompson, formerly with Evant and now Manhattan’s director of sales. ‘But it’s only since store-level ordering and replenishment have become more commonplace that interest in the technique has grown.’
Multi-echelon replenishment
Multi-echelon replenishment, too, has similarities with flowcasting and generates just as impressive savings. Alliance Boots, for example, has adopted the system across Europe, starting with the UK wholesale operation and its retail business in Norway. Preliminary estimates of savings based on cutting one day’s worth of safety stocks suggest a £30-plus reduction in inventory in the UK alone. Similar savings are emerging from the Norwegian operation.
While retail supply chains have become slick and sophisticated in recent years, for most chains the approach is still buyer-driven – a case of ‘you sell what we buy; not we buy what you sell’, as Dutch fashion retailer WE recently put it. In the flowcasting model the store-level demand forecast guides all else. Carried to its logical conclusion, if the forecast is right inventory matches demand precisely and there are no safety stocks left in the system. So where does that leave the seasonal clearance sale or those special promotions and deals, all driven in part by burgeoning stocks somewhere in the supply chain?
WE, which has implemented planning tools from i2 Technologies that have dramatically cut safety stocks, now has to buy-in special cut-price lines for its clearance sale.
