Demand forecasting 101
It is obvious that managing inventory stock for an e-Commerce business can represent a real headache, but nowadays, it is a very important part of any company. An inefficient use of warehousing space can cost a lot of money, and having too much inventory stock can also mean that products will end up becoming obsolete or damaged. But if you get your inventory stock levels right, you will save money and you will also increase trust in your brand. There is an interesting concept around the eCommerce market that plays with the idea of having a thoughtfully planned inventory, and it’s called demand forecasting. Demand forecasting is a process of anticipating the demand for the product and services of an organization in the future to optimize supply or execute decisions. This means that it is vital to plan ahead, which is certainly a very difficult task, but there are many ways to keep a healthy inventory using this concept. The following article will explore different tips for demand forecasting successfully your e-Commerce inventory.
Pay attention to seasonality
If you pay enough attention to your eCommerce inventory seasonality, that will definitely help you understand your sales turnaround. In short, there are periods of time during the year when businesses receive higher and lower volumes of customers orders. Every company goes through slumps while other points in the year are difficult for companies to keep up with demand. The most obvious example in North America is the fourth quarter of the year, which include the major holidays. Thanksgiving Day, Black Friday, Small Business Saturday and Cyber Monday tend to cause retailers serious headaches if they aren’t prepared to handle a massive spike in orders.
If your brand peaks around the same time as most others in the market, this may affect production lead times with your suppliers. Planning for longer delivery times from your suppliers or making sure that they prioritize your orders, will ensure you have a successful holiday season.
Historical Sales is a great place to start….if you have some
Forecasting your demand can be an easy job if you have the right data to start with, which means that you basically need some kind of historical sales data. Past sales helps you understand how the market is going to be functioning over the year, and if you combine this with your estimated growth, you will have a starting point to restock your products. For example, if last year you sold 100 units of a certain product, and your estimated growth is approximately calculated in 10%, 110 units of your product should be a good number to start restocking.
If you don’t have any historical sales data, either because you’re starting a new business or if you’re new in a certain category, there are still some useful tools at your disposal:
- Google trends: visiting Google Trends and typing in the product that you are planning to sell. Google Trends displays the interest of that keyword over time. You will see that there are months where people search the product more than others -for example, December-; you can leverage this information to stock the product right on time. You can also see if interest is on the rise or decline across different geographies and find some specific related terms that are rising in interest, even if they are not ranking at the top yet.
It’s not an exact science, but it will help you gain knowledge of the market and best of all, it’s free!
- Sales rankings and similar SKUs: Another great reference are public sales rankings for competitor SKUs. Although you may be launching a new item, surely there are competitors out there and you can compare their sales rankings with your own in the same category to get an idea of how much they sell.
If the item you are launching is similar to others you have launched in the past, you can use the existing article as a reference to start and then pay attention for the variances and adjust your expectations. This can also be done free of cost!
- Cognitive forecast: If the above methods are too manual for you and you are looking for something more sophisticated, AiHello offers a method specifically designed for this situation and it’s called Cognitive Forecast.
This method uses machine learning to find patterns and similarities among articles in a continuous and automated way. When there is no sales history, it strongly relies on article attributes and descriptions. As sales start happening, it will also look for patterns on the sales results. This algorithm is an incredible powerful tool for merchants of all sizes, particularly as they grow and have larger assortments to manage. The best part about it is that it will continue to improve over time and become more accurate through the powers of Artificial Intelligence.
Order enough to cover the time it takes you to get inventory from suppliers
In order to correctly explain this tip, it is necessary to unite two concepts: lead time and safety stock. The lead time is a basic concept that refers to the amount of time it takes from placing your order to receiving your product from the supplier. For example, let’s say that your brand is known for selling sport shoes from a european supplier. From the time you place an order with your shoes’ supplier, it takes 30 days for the product to arrive in your warehouse, approximately, from the moment the supplier selects your order, sends it to the cargo and it’s sent to your warehouse.
Those 30 days are your lead time, and this concept is really useful if combined with another calculation called “lead time demand”, which refers to the number of products that you need to keep in your inventory until the next shipment arrives. For example, let’s say that you sell 155 pairs of those european shoes a month, which means that you sell 5 pairs of shoes every day, on average (155/31=5). This will be your average daily usage of shoes. Because of the renowned shoes, that’s something you want to make sure you ALWAYS have in stock, so you multiply your average daily usage with your lead time (30×5=150), and that is the number of shoes that you should have on your inventory if there are no changes in the market.
Now, you want to know what happens with your stock if the market changes, and this is when the safety stock concept comes into play: safety stock refers to the amount of inventory a business needs to have to achieve a certain level of risk mitigation when it comes to stockouts. Safety stock provides a buffer of inventory for you to dip into when a counterproductive circumstance occur, and it is calculated as it follows:
Safety stock = (Maximum daily usage x Maximum lead time in days) – (average daily sales * average lead time in days).
In the aforementioned example, you already know the average daily sales number: 5 shoes, and you also know the average lead time in days: 30 days. Now you need to know the Maximum daily usage: think about a day in which sales have been on its maximum level, in this case, let’s say that in weekends, this store sells 10 shoes a day. Then, you need to know the maximum lead time. Well, let’s say that there are times -like winter holiday seasons- where the lead time have taken as long as 45 days.
Now that you have all the data to get your safety stock, you’d do the following:
Safety stock = (10 x 45) – (5 x 30) = 450 – 150 = 300.
Your ideal level of safety stock is 300 pairs of shoes.
And finally, a really useful formula that will let you know exactly when it’s time to place an order for a new shipment of products is called “reordering point”. It’s calculated as it follows:
Reordering Point = Lead time demand + Safety Stock
Using the last example, your reordering point could be calculated like this:
150 + 300 = 450. So once your stock hits 450 shoes, you will need to place a new order with your supplier. At 450 shoes, they’ll have enough to last them as they wait for new stock to arrive (150), while holding enough stock (300) as a buffer against an unexpected surge in demand, market changes or supply chain problems.
Consider having an integrated and centralized inventory management system
An inventory management system can help you manage your inventory stock more efficiently. Inventory management systems are designed to simplify the arduous processes that were mentioned before, like ensuring you are alerted when stock levels are low, or even processing returns from customers. It is designed to speed up ordinary tasks accurately, leaving you confident and with more time to focus on more creative tasks.
By following our advices, you’ll have a way to keep a healthy inventory. Forecasting your inventory is an accurate technique to plan ahead your orders and maintain a good amount of items in case of market changes, holidays and sales slumps; which will certainly help you improve your brand trust and save money. If you have any questions, don’t forget to ask for professional support for your e-Commerce products here at AiHello.