Ultra-Fast Deliveries For Top Customers Of Ecommerce

Customers buying something online always hope their deliveries will reach immediately. Top customers of an online retail store usually buy more than others, and at a certain time. Online retailers are looking at when their too customers buy, consistency of the time that a purchase is made, to activate deliveries, or even pre-place products at self collection lockers near customers' delivery locations.

Slotting Logic for New Products in Vertical Carousels

In cities like Singapore where land is expensive, automation is the way to derive more throughput. Retailers are using more putaway and picking machines like vertical carousels to maximize vertical space use. For some fast fashion products though, the popularity of products vary greatly throughout the week, by launches, and by seasons.

To minimize putaway and subsequent picking effort for new products meant to replenish products sold, a logic has to be in place to relate that new products are to be placed at the vertical carousel, and analysis done to locate this product with related products that are picked/sold together, that are at the same shelf. Not too different from market basket analysis done for up-selling, cross-selling and bundling analysis.

Auto Batch or by-Order Picking Logic (Base on Time Studies) In Picking Systems?

In most systems, picking of items are either done in waves/batch, or by invoice. Generally, when orders are relatively little, picking my invoice makes the most sense. As orders increase, picking my wave/batch minimizes pick distance, but one has to provide for shorting space to eventually split a pick into specific orders.

Most inital, or even mature, Ecommerce implementations involve using current warehouse or storage area for picking, alongside picking for B2B, or other B2C transactions. With high variety of products generally on a Ecommerce retail front, the long tail of consumers buying a few of a wide variety of products mean sometimes picking by orders, even for large operations, more feasible. The lines in the orders can be sorted so that the picker does not need to backtrack, and fulfill as much orders as possible.

There is even opportunity to look at time study specifics in picking to customize a system to automatically decide if batch or by order picking is more cost effective.

Dynamic Safety Stock and Reorder Point

Current safety stock and reorder points are relatively static. It assumes future sales with time is base on average historical sales, and vary with normal distribution.

In a competitive sales environment where substitutes are readily available at the right price, the assumption that future sales is base on historical average sales no longer holds. More and more companies are using forecasts for future demand to decide how much inventory to hold.

 

Future sales forecast, and the accuracy of sales forecast can be used to adjust safety stock and reorder points monthly. This because changes in average demand, variability of demand, and forecast accuracy affects safety stock levels, and in turn, affect reorder point. Base on demand characteristics, safety stock and reorder point can be updated monthly, or even weekly.

Unique Serial Number on URL in QR code on Packaging To Facilitate Packaging Tracking

Met a company that produces plastic bags for global retail gaints, and it got me thinking about how the humble plastic bag can be more useful. If each plastic bag has a pseudo-random serial number, complemented with a url like google docs, then one can set the plastic bag into part of a track and trace system for whichever contents it takes. It could be high-value medication that one wants to secure, luxury goods, or even a group of products with the same expiry that that a user wants to track together. It could leverage on Electronic Product Code - Information Systems (EPC IS) standards to enable a host of supply chain capabilities

One can link it to the point of use for pull supply chains to happen. Scanning the QR code also facilitates capturing of GPS location and timestamp for heatmaps and velocities to be analyzed.