As discussed in our last article on inventory planning and replenishment strategies there are essentially two approaches you can take to managing stock for the products that you sell:
- Fix the reorder quantity for each of your products
- Fix the times when you will reorder products
For many companies, especially small to mid sized companies, it’s much easier and more intuitive to go with the second approach, perhaps reordering stock from suppliers every month, every week or every couple of days.
Regardless of how often you order stock however, the main problem that you’re trying to solve when placing regular reorders is “how much stock do I need to buy (if any) for each of the products I sell?”. This can easily be done using demand forecasts and the new Replenishment module in SkuBrain.
Welcome to SkuBrain Replenishment!
To use SkuBrain Replenishment, all you need to do is select a product forecast and tell SkuBrain the lead times and current stock levels for the products that you sell:
SkuBrain takes care of the rest, automatically identifying any products that need reordering as well as any that are overstocked.
You’ll get a detailed stock plan from SkuBrain that you can view directly from SkuBrain or export to a CSV. The report includes specific reorder quantity recommendations and overstock quantities (that can be used to plan promotions and discounts):
SkuBrain tries to make things super simple and by default it adopts a single strategy (called a Universal Policy) for all of the inventory items that you carry.
Out of the box, the universal policy will assume that you order stock every 14 days, will try to avoid 80% of stock-outs and will consider an item overstocked if you are holding more than 200% of the inventory required to meet forecasted demand for that item.
However, you’ll notice an ABC switch in the replenishment settings that you can flip to switch to ABC Planning mode. Using ABC planning, you can split the products that you sell into three separate classes (A, B and C) that correspond to your best selling, next best (average) selling and worst selling products. You can then configure separate replenishment strategies for each product class.
For example, you might really want to avoid stock-outs on the best selling products and be happy to sacrifice service levels somewhat on your more obscure products in order to free up the cash to do that. In that case you could configure a service level of 95% for class A products, 80% for your class B products and maybe only 60% for your class C products.
Note also that you can classify products as Class A, Class B or Class C according to various different criteria. In the screen shot above products are being classified based on raw sales volume (i.e. the number of units sold). However it’s also possibile to have SkuBrain classify based on revenues (quantity × sales price) or profit (quantity × (sales price – cost price) ).
How exactly does this work?
It may not be obvious how SkuBrain calculates order quantities from the desired service levels (the chance to avoid a stock-out). When viewing forecasts in SkuBrain, in the top right you can see a a nerd switch. If you flip this switch then you’ll see some details on each of your forecasts, among which are some prediction intervals.
The prediction intervals give the statistical liklihood that sales will fall within a certain range. For example, there is a 95% probability that sales will be below the the
High 95 (95th percentile) in any given period and a an 80% chance that they will be below the
High 80 forecast. If you hold enough stock to meet the demand forecasted by the Hi95 prediction interval though, essentially you’re holding enough stock to avoid 95% of stock-outs, so these prediction intervals (and the many others that aren’t displayed but can easily be calculated) can be used to define service levels for the products you sell. Obviously, the higher the service level you strive to achive, the more stock you have to carry and products that have irregular sales are also more expensive to stock (although those products will probably be Class C products if you are using ABC planning with SkuBrain).
The rest is some fairly mundane math. If you’re really interested you can read about periodic review systems on the NCSU website. What the NCSU website refers to as
safety stock is calculated in SkuBrain by the prediction interval corresponding to whatever service level you have configured.
In the screen shots above, you can see we’re using a CSV file to upload data on current stock levels and, similarly, the resulting stock plan with buy/promote recommendataions can be exported to CSV format, which means you can use SkuBrain replenishment with any order management system under the sun.
However if you’ve configured SkuBrain to work with your order management software then SkuBrain can fetch stock levels directly from your ERP platform, avoiding the need to export and/or prepare intermediate CSV files. Similarly, the Stock Plan can be used either to create purchase orders in your ERP software or to push reorder points/quantities back into your ERP software (so that you can use the ERP software to generate these purchase orders).
That’s it for this week! We’ll no doubt eventually look at adding support for [fixed] economic order quantity replenishment as well but the focus for the remainder of the year will be on integrating SkuBrain with further ERP / Order management platforms and getting as many people on board for testing as possible, so please let know if you’re interested in using the SkuBrain beta, if you’ve got any ideas or feedback or even if you just want to catch up for a chat about how SkuBrain can help manage your supply chain.