Understanding Computer Algorithm Buying

TL;DR (Short Version)

Computers can make more efficient decisions by weighing all factors unemotionally to produce the best results. This can be especially important when dealing with products that have very high demand (like toilet paper during a pandemic). People still have a place in buying, but working with computers is what will create and execute the best buying strategies. To learn more how about K3S can help you create your company’s best buying strategy, click here To Get Started.

 

                           “Tell me and I forget. Teach me and I remember. Involve me and I learn.” – Benjamin Franklin

 

Computer Algorithms and Buying

A computer algorithm is able to take a look at many different scenarios very quickly. It then compares and contrasts projected outcomes to create a strategy that would be the most cost-effective while also meeting the requirements. Using the EOQ formula the computer was able to find the right balance of how often to order (affecting Order Cost) versus how much to have on hand (affecting Carrying Cost).

The first part of the computer’s strategy was to order every other month. This immediately reduced the Order Cost from $24.00 (ordering every month) to $12.00. This strategy, for this particular toilet paper product, was very effective. However, it did come with a price of larger orders.

To balance out the more infrequent ordering the computer needed to carry enough stock to handle two months worth of potential demand. This meant calculating an average and forecasting how much safety stock to carry. The safety stock needed to be enough so that spikes or dips in demand did not overwhelm the warehouse with inventory nor deplete too much to miss the service level target. 

One of the biggest advantages a computer algorithm has is that when inventory spikes or dips happen the computer will not react, or overreact, to these events and compromise the buying strategy. 

This delicate balance between Order Cost and Carrying Cost is evaluated and calculated every month to produce the smartest strategy for the current moment. This allowed the computer to meet the service level requirement while only spending $23.50 on our toilet paper.

 

Normal Strategies of People

When ordering products most people will start with one of two simple strategies. Either ‘I never want to run out’ or ‘I want to run lean and just barely meet my service target’. Both are valid strategies, to begin with, but when the demand comes in and the product begins to move people start reacting.

                    To quote Mike Tyson, “Everyone has a plan until they get punched in the mouth.” 

It is common for people to make reactionary decisions when purchasing and carrying inventory. This happens because of the external factors and pressures that come with being a buyer.

For example, a person may see that their inventory gets extremely low because of a spike, then chooses to make a large purchase in preparation for another spike that may or may not happen. They fear if they do not have enough inventory when that next spike of demand occurs they could miss their service targets. This one moment of increased demand can get ‘burned’ into their memory and for that one product they are constantly preparing for that next spike by carrying heavy inventory. They do not want to lose out on their sales and were probably reprimanded for missing out on sales by someone higher up in the company. 

Another extreme is they may see they are getting too heavy on inventory and decide not to order anything for one, or a few periods to thin out the large stock, and then run out. Trying to meet the service level requirements would have them constantly reacting to inventory demands, and always finding themselves behind the curve.

The buyer may also decide to run heavy on all inventory and never miss a sale. While on the sales side of the balance sheet this appears very profitable, on the warehouse/inventory side there would be an investment in inventory that is 50%-100% more than is needed. This extra inventory cost tends to be hidden because it is in “product” that is “just gonna sell anyway.” However, all inventory carries with it a cost to have and hold within a warehouse. To learn more about Carrying Cost you can read our Inventory Iceberg article.

These strategies and situations are played out for every single product at a distributor, not just toilet paper. Managing all of these effectively becomes an overwhelming task of how to efficiently and accurately purchase inventory for a distributor. A person by themselves will struggle to perform all of this analysis needed for every single product at every single supplier. This is why a computer algorithm is able to handle the task of forecasting how much to buy much more effectively.

People Are Necessary for Buying

The human aspect of buying is needed. Computers and algorithms simply cannot know everything about a company, the situation of a product, a supplier, or even the current state of the world (i.e. the pandemic in 2020). A computer is only as good as the data in front of it, and the suggested amount is only that, a suggestion. The ideal buying situations pair computers and people together to produce the best results leveraging the strengths of a computer with those of a person. Combining those together is, what K3S believes, the secret to purchasing the most effective.

Average Results of The Buying Game

We have given this game in paper form to individual companies in consulting, to auditoriums of people at conferences, and to classrooms wanting to learn more about forecasting and purchasing more effectively. If we were to give this to a room of 100 people the breakdowns would statistically be:

– On Service Level: Typically 50% of the room will meet the service level and 50% of the room will miss their service level target. Usually, this service level target is the main indicator for people that they ‘won the game’.

– On Total Cost: The average total cost for a person playing the game of those 100 people would be $33.00 with a deviation of $4.00.

In our experience the 50% of the room who meets their service level target (and is usually quite proud to do so) had to spend an average of 40% more than the computer to produce the same results.

Over the years of giving the game there have been a few people that beat the computer on cost and met the service level requirement. We estimate it is about 1 in 1,000 of everyone who has played the game on the first try. 

NOTE: This game uses a simple product with a perfect supplier that never makes mistakes, misses deliveries, or misses sending inventory. Also, the demand coming in is always accurate. A normal buyer purchases hundreds, thousands, or even hundreds of thousands of products. It is simply impractical to expect a person to be able to compete with a computer when dealing with that many products at scale. 

 

Thank You for Playing

We hope you have enjoyed our game and can appreciate the challenge of buying! 

 

Ready to Learn More?

To learn more about K3S (King III Solutions, Inc.) and how we help buying departments fill more orders with less inventory click the button below and fill out our form. You can also join our mailing list by registering below. We look forward to hearing from you!



Copyright © 2021 King III Solutions Inc.
K3S Logo