“I understand that we should replenish our points of sale with DDMRP for permanent products, but can you also do it for promotional products?” Our client asked us.
A year later, after multiple simulations based on sales history, and the implementation of a dozen promotional campaigns, the demonstration is there: it works very well, but the model needs to be adapted.
What is ADU?
ADU, average daily usage, is an estimate of daily consumption which is determined by either a historical average or by looking at future demand.
How to Calculate ADU
The challenge in the retail sector is to frequently replenish the points of sale in the most evenly distributed manner possible, and therefore to compute a daily ADU allowing to adjust the size of the buffers for each store.
There’s no secret, when it comes to promotional campaigns, you need forecasts. The tactic we have implemented is to base the sizing of store buffers on forecasts, before and during the first days of the promotion. After a few days we switch to 50% forecast and 50% historical since the beginning of the promotion. One week before the end we switch back to forecast. While in an industrial environment the CMJ is often smoothed over a long period, for example 12 weeks, we calculate it over a maximum of 14 days in this case.
Warning: the forecast is not established at the store level, it is carried out at an aggregated level, country or region, and broken down by store through relative weights to adapt the ADU of the day. Remember, the more detailed a forecast is, the less accurate it is, from our experience generating forecasts at the store level is not relevant.
How do you size the buffers?
The buffers are mainly conditioned by four elements: how often the store is replenished from the hub, the transport time, the variability of demand, and the amount of visual merchandising.
Depending on the frequency and the lead time we size the red zone base, the green zone and the yellow zone – using the interval between deliveries. We size the red safety zone according to ADU thresholds. For example, for an ADU less than 0.2 we have a high variability, and a low variability for an ADU greater than 1.
Let’s not forget the amount of “visual merchandising” – in other words, the quantity needed on the displays so that it doesn’t look empty, and that it makes you want to buy. We include it, in quantity, in the red zone. For items with low turnover, this quantity is enough to size the red zone, otherwise we add a dynamic red zone.
How to manage deployment?
Deployment to the stores is done automatically at night. Beforehand, during the day, the planners adapt by exception some buffers, and process the execution alerts. They focus solely on monitoring the operating model.
In a situation of constrained supply, the Replenishment+ stock deployment module ensures the inventory deployment using a fair share per store.
What are the benefits?
Replenishment of stores via DDMRP brought new visibility to the teams. How does the stock evolve in relation to the cumulative red/yellow/green zones of an item across stores? What is the impact of the replenishment frequency? What is the impact of visual merchandising? What is the effectiveness of the buffers? Replenishment+’s business intelligence functionalities were used to establish an adapted dashboard.
From the flow managers’ point of view, the model is quickly understood, there is no black box effect, and decisions are made in time thanks to execution alerts.
Beyond the stores replenishment, the process makes it possible upstream to regulate consumption at the distribution centers, to also transition them to pull flow, and thus to go all the way back to the factories and suppliers, for much greater agility from one end of the chain to the other.
If you are fortunate enough to be able to pace your supply chain in pull flow from point of sale information, don’t hesitate, DDMRP is the ideal way to do this!