When your stock runs into thousands of SKUs (Stock Keeping Units) you often resort to blanket business rules and statistical methods that work for some items but not for others – intermittent demand items for example. One area of inventory management that affects many businesses we work with is seasonality.
To be really on top of your stock and hit that sweet spot of not holding too much inventory, and at the same time being able to meet demand, fulfil customer satisfaction levels and not lose out to the competition, you need to be constantly reviewing and amending your inventory accordingly.
One measure used by a number of our clients is to calculate “weeks’ cover” by taking the historical sales quantities from a previous relevant sales period and deriving a weekly sales average over that sales period – which you then divide into your Stock on Hand figure. But if you have seasonality in your product range you need to be very careful when choosing the historical sales period upon which you are basing your weeks’ cover calculation.
Whilst training some users in the Market Gardening sector recently I noticed that it was hugely more complex when it came to varieties of Summer Bedding Plants or hedging. Even within a relatively small category there were variations of seasonality that could make a huge difference to determining stock availability based only on weeks’ cover for the category as a whole, using the same sales period for all products in that category as the basis for the calculation.
When products are so seasonal and perishable (must be planted after a certain date due to frost or before another date for best root formation) it can make huge differences to the bottom line if you have to throw out stocks of pot bound plants or packets of spring bulbs.
WHERE DO YOU START?
Well our advice is that you begin by putting the most effort into setting up key reports that analyse as accurately as possible the Core Products that give you the most profit year on year. Typically we know from experience that the products that give you say 50% of your profit might only comprise 10 or 20 % of your product list. If you can get this part right, then the less profitable and possibly the biggest part of your inventory can be analysed with more generic methods.
USING THE EXPERTS IN YOUR BUSINESS
With seasonal products especially it makes a lot of sense to call upon the product experts that you already have within your organisation – the people who have experience in managing product categories and meeting the demand for products from your customers in their areas of expertise. This knowledge, when used in conjunction with some automated techniques that have been developed for breaking down large inventory lists into groupings, is crucial for a business when beginning to design automated routines.
ABC ANALYSIS ON PROFITABILITY
Using a technique such as “ABC analysis on profitability” as a starting point, you can divide your data into key products and the rest (there are a number of ways of doing this – ABC analysis on profitability is just one). ABC analysis looks at sales history of your products over a defined period of time, sorts them high to low on profit delivered and categorises the products into groupings or bands A, B, C, D, Z and New. These bands come from predefined thresholds of the % of total profit each band has delivered in that sales period – often 12 – 18 months. So for example, the top 20% of profit could have been delivered by 10 products (Band A), the next 30 % by 200 products (Band B) and so on. Z Class delivered no profit in the historical sales period – typically this could be one of the biggest classes, and New would be the latest products with little history. It is important that the results of a first pass at this automated routine should be sense checked by someone (or a number of people) with knowledge of their respective category(s).
Once a core range has been defined, then within only that core range start focusing on the data that is seasonal currently or will become so – i.e. well in advance of the lead time. You can then set up rules for lead times and calculations of weeks’ cover based on the correct seasonal historical sales period so that your averages are based on a like-for-like period.
This ABC analysis itself could be set up so that it runs against seasons if your business is extremely seasonal.
You can also look at the average number of sales units of a particular product by transaction to get a feel for e.g. what are the number of plants purchased for an average hedge.
By using your in-house experts and walking through the steps of how a category manager would manually proceed to work out seasonal stock requirements you can incorporate these steps as business rules in your reports – and these will probably be distinct for the different types of products you sell.
So by putting in the effort at the point of grouping your data into Core Products and within those, the seasonal products, you can begin to define much more accurate and focused reporting which, once set up, can be run automatically and used regularly at the appropriate time of your sales year.
Of course there may be ongoing variations and tweaks to be made, depending on local knowledge about events and external influences such as weather or competition, but you will have a foundation on which to build. You can incorporate other data sets into your analytics (using an integration tool such as the Diver BI Visual Integrator within Workbench) that will help you to build a more complete seasonal picture. This will allow you to look for correlations in events, weather patterns and sales peaks or troughs that your internal experts might be able to pick out as causations for sales, thereby creating further reliable rules for future patterning.
Even with specialist forecasting software that takes into account seasonality and slow moving products it always takes human involvement and expert product knowledge to manage the first pass of the forecasting algorithms and tweak the results regularly.
EXPERT REPORTING
The key aim of this type of report is to create analysis based on both the type of number crunching that only computers are capable of; and incorporate the expertise, the sense checking and context definitions that only someone with experience of your business has the knowledge to include. This type of “Expert Reporting” is the best way to achieve winning results.
We have seen that where companies have made this effort they are reaping the rewards with increased sales.
See the recent related blog on the importance of understanding the context of data by Kathy Sucich from DI. http://www.dimins.com/blog/2017/08/07/the-downside-of-data-science/
About Debbie Lonsdale
Debbie Lonsdale has been working with the Diver Solution as a BI Consultant at Dynamic Business Informatics since 2008. Her previous experience includes computer programming, analytical and technical roles, team management, account management, sales and marketing in a variety of market sectors, including the Travel Industry and Distribution. She combines this experience as an all round ICT professional in the BI sector.