Previous Article: 10 years already!
Next Article: The Language of Machines

Nov 08, 2021

Yes, predicting your sales is possible!

by Hugues Foltz Executive vice-president

Artificial intelligence

Mountain bike is a passion of mine. One weekend, I unfortunately damaged my derailleur – a tragedy for someone like me who rides several times a week and who knows that derailleurs have become a rarity in any bike shops whatsoever.

If there is one thing we learned from the COVID-19 pandemic, it is the sudden scarcity of an incredible number of consumer goods. Equipment for anything close to a hobby or sport became unavailable in stores. And It’s not just the substantial demand for certain products that explains shortages of all kinds. The breakdown of supply chains during the crisis and the prolonged shutdown of factories also had a significant impact on many industries. Seeing the problem at hand, one can surely ask: would it have been possible to predict this peak in demand for certain sectors?

Predicting sales is the dream of countless entrepreneurs and CEOs. Wouldn’t life be easier with a crystal ball to foretell such things for us? By knowing the demand in advance, we can anticipate the correct number of people to ensure production, control the amount of raw material needed, manage stock in the warehouse and, above all, we can produce the right thing. In an increasingly competitive business environment where profit margins are ever thinner, the costs associated with maintaining inventory, managing overtime hours or not being able to deliver orders can often make all the difference at the end of the year.

The availability of extensive databases of all kinds is multiplying. The variables that we can correlate with the sales history of companies allow us to shed light on trends and lead to AI models that would surprise many entrepreneurs with their accuracy. Unfortunately, it seems managers and sales teams need to put a big dose of sentiment into the annual sales assessment they believe they can achieve the following year…

Dealing with the unexpected

Many companies, however, have already made the leap towards AI. One of them, Stove Builder International (SBI), has accepted to reveal their ongoing project. With 5 different brands of stoves, more than 240 products and several stores across five continents, SBI almost exclusively uses the sales prediction of its AI model to plan production and growth, one year ahead of schedule. The impact of such a prediction goes far beyond fulfilling all the orders that will be received. When you know the "future sales" variable of a business, it allows you to do strategic planning from a whole different perspective!

As with anything AI-related, the accuracy of a prediction is always key. In general, it is difficult for a sales team to forecast sales one year in advance with a precision rate over 60%. SBI’s team is now able to forecast their annual sales with 85% accuracy. That’s what I call a game-changing advantage! If SBI can do it, why not your own company?

Imagine growing product at the right time because you were able to predict that the vegetables arriving from Florida would be late. This gives you control over when to put your products on the shelves with the confidence that buyers will pay the price you ask. Better yet, you can make predictions based on consumer habits, the time of year and a host of other variables translating the optimal amount to produce to minimize waste.

I believe that the power of today’s artificial intelligence models represents so much more than an opportunity for companies to make more money. It is also a great way to reduce waste in the food industry and the overproduction of goods with a very short life cycle, such as clothing.

Let's use this great technology to rethink our production, and thereby have a positive impact on the planet!