Are you handling the Supply Chain or Inventory of your retail business, and are looking for ways to optimize your decisions ? Here’s how data analytics can help you.
Keep a record of your historical inventory
Why Maintaining Inventory History will improve your forecast accuracy and avoid opportunity losses due to Out of Stock (OOS).
What/How: Save historical records of opening and closing stock (ideally, daily) to calculate historic out-of-stock (OOS) occurrence and lost sale units. In the context of a sales forecast, you can adjust the historical sales by including lost sales units. Including these past opportunity losses will provide a more accurate prediction for future sales.
Keep track of your procurement, order and inventory adjustments
Why Adjustments are the source of error in Planned versus Actual sales, and if not accounted for are practically untraceable.
What There are 2 types of adjustments:
— Analytical (automatic) driven adjustments: rounding to supplier batch size, Minimum Order Quantity (MOQ), or Full Truck Load (FTL).
— Manual: Risk Takers/Account managers tend to deviate from forecast at an operational decision level.
How Avoid analytics adjustments by breaking down forecasts for demand and supply planning such that it is a Highest Common Factor (HCF) of MOQ/Batch Size etc. at each stage of Supply Chain. In case HCF is too high you will need to negotiate (say on the supply side for MOQ) or forecast at a higher aggregated level. Manual adjustments should be captured using predicted / planned flow of material vs actual flow. Allowance of deviation should be allowed but needs to be controlled with the help of policies.
Use the Product Life Cycle (PLC) to adapt your forecast
Why Too often companies ignore PLC while making demand forecasts for their products. This leads to erroneous estimations and inventory imbalance.
What The most important parameters to look for are 1) the economic indicators such as Disposable Income per Capita and 2) competitor activities such product launches and promotions.
How For market sensitive products such as Smartphones, frequently check and account for the effect of similar or overlapping product launches from competitors, any technology change (4G to 5G) , economic change (Covid 19) etc. These factors can change your products PLC. For example if the maturity stage reaches early it would be a good idea to adjust your forecast and see if the decline stage would also reach early. We recommend using regression analytics or time series forecasts (such as vector autoregression) for optimal results.
Use analytics to plan your shelf space and increase sales
Why Shelf visibility is a big factor in product sales. Moreover, shelf space is limited and precious. You should thus use analytics to plan shelf allocation precisely according to the product demand.
What/How Observe past sales data to place high demand products on priority shelves with optimal visibility. Set up automatic notifications to trigger replenishments. Optimal On-Shelf quantity differs by product and retail store. Important considerations for deciding on-shelf quantity is first replenishment lead time and second average distance between the shelf and customer so that product is adequately visible for making a purchase decision.
Prevent FTL policy from affecting your replenishment needs by optimizing order design
Why FTL policy prevents you from replenishing until the policy is met. Either you face OOS or incur extra cost on adhoc replenishment.
What/How Think of logistics as a service and optimize on multiple parameters in each run. For example
- Identify the probability of FTL given different frequencies of replenishment.
- Enable returns with replenishment to maximize truck utilization in each run.
- Plan route for multiple pickups and drop in forward as well as backward journey.
Specialized route optimization algorithms can help you to plan these services and best utilize your fleet capacity.
Manage your peak hours with overlapping shifts.
Why A good shift management will help you avoid overstaffing during hours of peak demand, which will save you costs.
What/How If peak hours are shorter than the number of hours in a shift (for example, 4 hours of peak sales in evening shift) you readjust the shifts such that they overlap during peak demand, ensuring sufficient manpower without additional staffing. Redistribute workload and tasks during these hours to ensure accountability.
GSC is a data analytics and software development consulting startup. We apply Machine Learning on historic data to help you make better business decisions. Interested in learning more ? Contact us.