Home / Soft4Inventory Blog / Critical Thoughts About Forecasting as an Inventory Management Tool

Critical Thoughts About Forecasting as an Inventory Management Tool

Author Simona Skorupskaite  Simona
LinkedIn logo

– The average reading time for this post is 4 minutes 30 seconds –

Monitoring SKU quantities and effective replenishment estimations are essential to keep the right amount of stock on hand. There is a long list of widely usable inventory-management systems that are integrated with forecasting tools and usually applied to tame demand volatility. Forecasting is said to be powerful, because it provides an intelligent and model-based ‘glance’ into the future.

It seems to me that there is a common belief today that improving forecasting accuracy should definitely improve inventory levels. Very often, I am faced with a business tendency to rely on sophisticated forecasting models. So forecasting is hot, but have you ever attempted to explain what it is, and how the predictions are made?  Let me introduce to you some critical thoughts about forecasting as a tool to prevent stockouts and improve inventory levels.

  1. “It is often said there are two types of forecasts … lucky and wrong!!!” (Control magazine). The term “forecasting” is usually confused with planning, but it can be described quite simply – it is a model of historically monitored data. The model can be expanded into the future, thus giving the predictions. This is fine, but there is always some level of confidence within every model, which means the evaluated predictions have a non-zero probability of being wrong.
  2. Nature of demand instability. Not surprisingly, classical forecasting algorithms are based on minimising residuals. However, this does not mean that the model explains the behaviour of demand variability. In addition, as a result minimal residuals do not guarantee that past events define the future, and vice versa; actual demand may not be related with historical tendencies. Indeed, I think there is a lack of understanding about what is the real demand potential in supply strategies.
  3. Data restrictions. Since predictions are closely related to historical data, take into account that there is no way to model the required quantity of incoming goods with inadequate historical information. Extreme events are also part of your business. However, forecasting is not tuned to include extremes, as such events are treated as outliers.
  4. When is it enough? Usually items tend to be segmented by seasonality, lifecycle spans, promotional scope, coherent specifics (demand of one item influences/generates demand for the other item, e.g. printing machine and toner) and many more factors. Since there is no unique forecasting model to describe various fluctuations, the systems use many of them in order to support different combinations of factors. From a theoretical view, data overfitting is highly expected. Meanwhile, as regards your business opportunities, the inventory management software system is not flexible enough to adjust to an increased number of segments. So you always find yourself seeking new models in terms of quantity and quality.
  5. Integration into inventory strategies. There are many vital reasons your business can collapse due to stockout: some changes in lead times, loss of a supplier or bulk discounts. Your inventory works as a buffer to absorb these interruptions when the risk of a stockout arises. This is why it is very important that inventory strategies are properly designed. Lack of understanding of a planning concept and inventory weight on your business finance should influence seeking for a structured inventory strategy.

Let’s have a simple thought experiment. Say that you have a neat little ice-cream business.

Monday. It is really warm outside, so you have quite a few customers. It is going to be even warmer tomorrow. Forecasting tells you that (based on historical data) the hotter it gets the more customers will show up. You prepare a large amount of ice cream and wait.

Tuesday. It is really hot outside and you have a lot of customers. It is going to be even hotter tomorrow! Once again, forecasting tells you that the demand will increase. You prepare an extreme amount of ice cream and wait.

Wednesday. It is blazing hot outside and… Nobody leaves home. Even though the forecasting projected a clear increase in demand, there is none. Your stock goes bad and you are out of business.

How could that happen? …

A bit of an exaggeration, but you get the essence. Considering this example, it can be concluded that forecasting is mostly restricted by unpredictable errors that can lead to stockouts or frozen capital. There are inventory management software solutions on the market that use dynamic buffer technology to set and adjust appropriate stock levels for each item in each location, using inventory-state levels rather than forecasting (customers keep asking me: could we set the inventory levels for families of items instead of single items. “Oh, we have hundreds of thousands of items, it’s going to be a nightmare to analyse them all!”. The answer is “no”. Why? Read here. And it will not become a nightmare. Our customers spend 20 to 40 minutes per day adjusting their inventory levels).

Use our Profit Calculator and find out your how much you can profit using Inventory Management software based on Pull approach!
Use our Profit Calculator and find out your how much you can profit using Inventory Management software based on Pull approach!

The “dynamic buffer” methodology incorporates current inventory levels as a signal instead of a forecasted demand. Dynamic buffer also allows evaluating inventory demand in response to lead-time changes, bulk sales or loss of suppliers. Still, to change something that has become a commodity – I mean a forecasting-based approach in this case – is very hard. Why? Because people do not like changing, even if they know it’s for the better. As was said in a great book by Switch – “How to Change Things when Change is Hard”, “The rational mind wants a great beach body; the emotional mind wants that Oreo cookie”. Maybe calculating how much more you could save or earn will change your mind. Let’s give it a try!


Check for more insights on forecasting and your inventory management:


http://www.forbes.com/sites/loracecere/2015/11/29/does-better-forecasting-improve-inventory-why-i-dont-think-so-anymore/

http://www.forbes.com/2010/02/11/why-forecasts-fail-leadership-managing-mitsloan.html

http://www.smh.com.au/business/comment-and-analysis/its-a-rule-forecasts-are-always-wrong-20151215-glnyz5.html

 

SOFT4, as a provider of software solution for efficient inventory management, is constantly releasing blog articles about how to meet and cope with the trends, tendencies, and challenges, and give suggestions as to how to react to the changing environment in order to stay competitive.

Subscribe to our monthly newsletter and get fresh insights for efficient your inventory management!

Become an Inventory Management expert!
Join hundreds of successful Inventory Managers, who get valuable information on stock management and optimisation straight to their inbox.
Contact or visit us

Business Development Manager: Mrs. Ugne Kontare
Mobile: +370 682 55216 | E-mail: ugne@soft4.eu

UK: +44 1732 617970
Netherlands: +31 35 7510 110
Rest of Europe & International: +370 682 55216
Canada and the USA: 001 844 341 5512 (toll free)
Email: ask4soft@soft4.eu