“Things change quickly. You have to adapt,” says Toby Brzoznowski, co-founder and chief strategy officer of the supply chain management software company LLamasoft. “The companies that aren’t doing this well are being left behind.”
LLamasoft concentrates on the data analytics and design aspect of supply chain management, helping clients use data to become more agile in a shifting landscape.
Founded a little over 15 years ago, the Ann Arbor, Michigan-based company currently has around 700 customers globally. Many are Fortune 500 companies with a billion dollars or more in revenue, Brzoznowski says. They represent a wide range of industries that include pharmaceuticals, oil and gas, electronics, automotive, and consumer goods.
Recently we caught up with Brzoznowski to find out how supply chain management is transforming in the digital age, the difference analytics can make for global companies, and his advice for how to avoid getting left behind.
What are the biggest challenges your clients face around supply chain data?
There is no shortage of data, but if they don’t have a good centralized data model of their end-to-end supply chain, they end up having siloed information. They can’t test trade-offs across their entire supply chain or across competing objectives.
Lowest cost can be a key objective. Optimized service is a second, and a third is mitigating risk. You have people within the organization whose job may be sustainability — reducing emissions and lowering energy use — but if they have to compete against any of those other three objectives at the boardroom level, it’s very difficult to win.
We’re finding that making the supply chain more efficient is having a secondary effect of improving sustainability.
Are there common mistakes you see?
There are too many decisions made on gut feeling and anecdotes. The more those decisions are replaced with a data-driven approach that allows new available analytics to take over, the more you can drive smarter decision-making.
Supply chain leaders are risk-averse, for the most part. They aren’t going to make a decision where they don’t already know the outcome. Historically that meant very conservative decisions. Now they have the ability to test decisions. They can predict the financial performance, the service metrics, the effect on their sustainability metrics.
What best practices do you recommend?
Execution and planning are the two pillars of supply chain management. In the last few years, there has been the emergence of a third pillar: design.
Execution asks, “How do I run my supply chain?” Planning is, “How do I make that supply chain more efficient?” Design answers the question, “Is this even the right supply chain?” This is where data science comes in. It can take all of the execution-level transactions and bring them together into a digital model.
Companies are creating centers of excellence with centralized teams that look holistically at the supply chain across the globe. They can question the status quo and the policies that the company uses — and try new things.
Could you share a client example?
A large grocer with operations throughout the United States was considering adding a new regional distribution center because they were running out of warehouse capacity. They wanted to identify the best location for a new site that took into account all the inbound product flows from their many suppliers as well as bracket pricing — stepped discounts based on the volume of items purchased.
They used our software to create a digital model of their supply chain, which enabled the company to evaluate millions of combinations of different product flow-paths, inventory stocking policies, purchasing volumes, and facility locations.
What did they learn from that?
The model showed that they actually didn’t need to spend tens of millions of dollars opening a new facility — their capacity issues could be solved by implementing a new inbound sourcing strategy and a new flow of goods.
For their top 20 suppliers, which made up nearly 80% of their overall volume, they identified a single distribution site to serve as an inbound consolidation center and central purchaser. This enabled them to make significantly larger purchases, reducing their per-unit price. Then they could push smaller inventory amounts to the other regional DCs along with other existing inter-facility transfers, reducing large bulk purchases at each facility, and in turn significantly reducing inventory.
The result was a nearly 10% overall supply chain cost reduction and the avoidance of a significant one-time facility spend. This new system provided an added benefit of an overall reduction in miles traveled, which reduced carbon emissions. It shows that traditional spreadsheet analysis can’t possibly evaluate all the options.
What happens when an organization has trouble getting supply chain information, like traceability data for a historically problematic raw material?
That’s becoming more of a challenge. There are companies realizing that the risk outweighs the reward, and they will go to suppliers that are more forthcoming with information.
Ultimately it’s a design and sourcing decision. Can I use an alternative material if I can’t get enough data? If I can’t use an alternative material, is there an alternative supplier that’s more reputable? And if neither of those exists, is this a market that I truly want to be in?
Do you see leaders growing more comfortable having machines do the thinking for supply chain management?
Absolutely. It is already part of everyday life, from sourcing and replenishment to what type of service to promise for an online order. Many of these decisions are already being made with human intervention. You’re going to see more of those examples over time.
How do you see supply chain management headed in the future?
Where it will go next is into areas of AI and machine learning. So you can learn from decisions, create policies, and wherever possible leverage artificial intelligence to make those decisions across millions of different variations faster and more effectively.
