Subject Area: Monte Carlo Simulation
An interview on monte carlo simulation and inventory optimization with Michael Worden, CEO of Right Sized Inventory
Software and techniques for inventory optimization have evolved over many years.
While this sounds like good news, the real issue is that the approach used in software is not always clear or transparent.
Recently I interviewed the CEO of Right Sized Inventory, Michael Worden, about Monte Carlo Simulation: Inventory Optimization, Spreadsheets, and Black Box Solutions.
The following is a transcript.
Phillip: Hello Michael. RSI, as Right Sized Inventory is sometimes called, was launched with a very specific idea in mind, but before we get to that, perhaps you can tell us a little about the company.
Michael: Thanks Phillip. RSI was launched in 2014 as a collaboration between Dave Cook and David McPhetrige, who is sometimes is referred to as McP. Their goal was to turn McP’s patented inventory optimization technology into a product to help thousands of small to large sized businesses. To do that they joined forces with an investor Chris Koski, and a software developer Steve Cook. Together these four individuals developed the software and launched the company.
Phillip: Ok, so you’ve been around for roughly six years and I assume the company has grown significantly.
Michael: That’s right. Today we serve companies of all sizes, across a vast range of industries. Many of our clients are household names, like, Anheuser Busch, Advanced Energy, Cooper Tires, Valvoline, Moen, and 3M.
Phillip: And you’ve grown beyond the original staff of four.
Michael: Yes, of course. These days our headquarters is located in New Mexico (USA) and we have team members distributed around North America. Beyond our direct staff, we have several strategic partnerships and resellers around the globe. We are also working with select companies who offer unique supply chain platforms to embed our patented technology into their offering.
Phillip: Great. In two sentences, tell me what RSI does.
Michael: Put simply, we take the guesswork out of determining the minimal amount of inventory required for a business to achieve defined service levels; thus, eliminating inventory imbalances and enabling them to quantify risk. We do this utilizing real-world simulations of their supply chain environment.
Phillip: Maybe you could expand on that?
Michael: Basically, we simulate the supply chain for any given item in any given location. We run the simulation for a defined period, let’s say for 90 days, but it can be any length of time. During the simulation, daily inventory goes up due to receipts and goes down due to demand or any other valid consumption. We then iterate the simulation using Monte Carlo methodology to find the optimal answer to this question: What is the minimum inventory level that has the highest probability of meeting the required service level?
Phillip: You mentioned Monte Carlo Methodology. In my book, Smart Inventory Solutions, I defined Monte Carlo simulation as ‘an analytical technique that uses random numbers as input variables and applies them to a known function to identify possible outcomes.’ Is that what you meant?
Michael: At a high-level, yes. Monte Carlo simulations build models of possible results by substituting a range of values, in other words a probability distribution, for any factor that has inherent uncertainty. It then calculates results over and over, each time using a different set of random values from the probability functions. Depending upon the number of uncertainties and the ranges specified for them, a Monte Carlo simulation could involve thousands or tens of thousands of recalculations before it is complete. Monte Carlo simulation produces distributions of possible correct outcome values.
Phillip: That’s a fairly general description for Monte Carlo simulations, so how does RSI apply this approach?
Michael: In the case of RSI, our probability functions are built for both the demand pattern and the replenishment supply chain. We vary demand and supply information on each day of the simulation utilizing values from each of these probability functions. Values are represented based on probability, so a demand value closer to average demand for the item will be represented more often than a value in either tail of the distribution. Ultimately, this method ensures the answers are robust as they have been ‘stress tested’ through many simulation scenarios to ensure the answer has the highest probability of being the right answer.
Phillip: That’s a lot to explain to people who may not be well versed in statistical analysis.
Michael: It is, but at RSI, we of believe our science and methodology is worthy of trust. So, we decided the right approach is to explain to people exactly how we do what we do. We aim to avoid the “black box” approach where companies say, “just trust us”.
Phillip: But surely, some people attempt to re-create your approach on their own.
Michael: Of course, but typically those individuals or companies, contact us after they realize their mind-boggling spreadsheets don’t deliver results. Our technology and methodology are patented, involving decades of expert work to develop and validate a capability that extends well beyond the scope and capabilities of any spreadsheet.
If I may tell you a short story.
In the earliest days of RSI, Dave Cook met with a key influencer who, in a not-so-encouraging moment, said:
“We’ve found that inventory optimization solutions as a whole have under-performed to expectations; this seems to be primarily due to a lack of comfort or trust in ‘black box’ solutions where the customer can’t clearly understand or explain how the answer was derived.”
Phillip: Wow, that’s tough.
Michael: Yes, but in all sincerity, Dave really appreciated that candor and guidance. The assessment was spot-on! Since that time, as predicted, prospective clients have voiced concern along these lines. Usually this concern goes something like this:
“If anything goes wrong, I’ll be expected to explain the error and planned resolution. I can’t point the finger at the ‘really sophisticated software’ or we’re both done!”
That’s why saying “just trust us” would miss the mark.
Phillip: That’s refreshing. One of the issues with suppliers of forecasting software is it usually requires a large upfront investment, even before the solution is properly proven within the company. One of the things I find most interesting about RSI is that your solution is scalable. Was that a deliberate strategy?
Michael: Yes, RSI was kept “skinny” on purpose. This strategy reduces our development overhead, which in turn keeps costs down, making it affordable to almost every company in the market. It also means companies can justifiably access our solution regardless of how many inventory items they are managing because our pricing scales accordingly.
Phillip: But your competitors offer a broader platform.
Michael: They do, but the features offered are either beyond the scope of RSI or of diminishing value and may only really be there to justify their pricing. Even if the additional features bring value to the client, RSI’s technology does not displace that investment, we enhance it. At RSI we say, why pay for more than you really need? Our feature set is narrow and that’s why we have gotten so good at it.
Phillip: Good point. Before we close, do you have any final comments?
Michael: Phillip, I will leave you with a few final thoughts regarding Right Sized Inventory: First, the market issue we solve is simple. For any company that carries stock, we eliminate inventory imbalances that introduce cost and risk to their business. Second, our founder’s mission is to provide the best inventory-optimization tool on the market that nearly every company can afford, and I am proud to say we are realizing that vision. Finally, I want to thank you for the opportunity to speak about Right Sized Inventory today. Many thanks.
Phillip: Thanks for your time.
Michael: Thank you Phillip.
Posted by: Phillip Slater