It’s Not A Supply Chain Anymore, It’s a Supply Ecosystem!

As Bob clearly points out in our joint series on Supply Chain Matters on why legacy sourcing and (supply chain) planning solutions can’t handle today’s supply chain challenges (and why direct sourcing needs to be supply chain aware), Supply Chains Have Evolved to Demand and Supply Ecosystem Networks.

They are not a simple linear chain anymore. They aren’t even a simple tree where a supplier uses sub-tier suppliers for parts that uses sub-tier suppliers for components that uses sub-tier suppliers for raw materials. A tier 1 supplier could be using a tier 2 supplier that is used by other tier 2 suppliers and a tier 3 supplier could be using a tier 1 supplier for products for manufacturing. Then you have intermediate assemblers and distributors that pull parts, components and materials in, assemble some, and package others in a bundle for resale. Maybe you are selling your product to a tier 3 supplier and don’t even know it. It’s a many-to-many graph. and a very convoluted network that can only be described as a complex supply ecosystem.

It’s a complex network that requires the orchestration of strategic and tactical product sourcing with supply management teams who might be participants in the customer fulfillment process, leaders of the sourcing process, managers of the logistic network, or all three. All depending on what is needed, when, where, and why.

If you only look at one need, from one perspective, at one point in time, you miss the fact that any decision not only impacts every other internal organizational unit in the supply chain ecosystem (which includes sourcing, procurement, logistics, supply chain, R&D, production, and operations), but has network wide impacts, implications, and response.

If you change the supplier for a part during a sourcing process and award a new contract, that affects Procurement since they have to update their catalog and reorder systems; Supply Chain as they need to have the appropriate network in place for cross-docking, temporary warehousing, and storage; logistics as they may need to onboard a new carrier to pick up from the local factory; Production as they have to confirm the part will work across all product lines the part being replaced was used in (and, if not, you may have to retain the current supplier at a much smaller volume or find a replacement); R&D as they will have to confirm the part is okay for all the products they are developing; and operations needs to ensure it is categorized and tracked properly so they understand the data shifts during their analysis and cash flow forecasting.

When you change the supplier, that could have a negative impact on the former incumbent who might have been continuously allocating 40% of their capacity to your business with no quick way to recover that (because they had to cool their pipeline to support you), especially if that part was 75% of your business. (And this could have a negative impact on you if you are relying on them for other parts.) Chances are they’ll have to do layoffs in the short term, and that will impact their OTD and quality on your remaining contracts.

When you change the carrier, they need to reallocate the driver. Probably not a big deal, but if they were subcontracting to a Mom & Pop Trucking Co. for a route they didn’t normally do, that’s a big deal to that Mom & Pop Trucking Co. that needs to find a new regular route fast. And it’s a big deal to you if cancelling that route drops you below the commitment you made (and you lose your discounts and special rates you spent weeks negotiating).

The impact on all of these departments and gears in the supply chain machine have to be considered, especially as those gears turn back towards you. To make the best decision, you truly need to do an integrated analysis across multiple levels of planning (long-term, mid-term, and short-term).

Moreover, this analysis needs to be done in near real-time as businesses need to be able to quickly pivot to changing demand or supply balancing requirements in today’s dynamic global marketplace (brought on by pandemics, border closings, canal closures and slowdowns, sanctions, economic upheavals, wars, and trade wars). This requires an integrated network view across business departments, views, and timeframes.

And, as we have said before, it requires that business finally Think Different, and stop reverting to spreadsheets as a means to attempt to span non-integrated internal and external information streams. But considering that Excel is still every analysts favourite tool and they all want to be King of the Spreadsheets, who knows when the shift to modern, integrated, analytics will finally happen.

Direct Sourcing Should Be Part of Supply Chain Management …

… but it’s not, and the reality is that none of the Supply Chain Management solutions (which are often still in the middle ages of technology) are anywhere close to even considering direct sourcing, so you can guess how close they are to being able to support it.

This means that, in order for us to get to the point where we have a direct sourcing solution that works, integration will be key. Furthermore, the integration must be as native as possible between the direct sourcing and supply chain planning and management solutions, with full, real-time, data exchange in compatible models.

This is because, as pointed out in Part 4 of our direct sourcing series with Bob Ferrari of Supply Chain Matters on how integration is key, without native integration, it’s hard to react and execute effectively when:

  • consumer demand suddenly spikes or flatlines
  • supply lines (or a supplier) suddenly become unavailable
  • costs significantly change over night

Only with direct integration to the supply chain visibility, planning, and execution systems will you:

  • see the spike when it happens, not when it’s too late to even attempt to do anything about it
  • see the delay in a shipment that could indicate a supply chain disruption and see the alternate supply lines or suppliers that could be rapidly be routed to or ordered from
  • see what your alternatives are, what they would cost, and how long it would take to bring them online

… and begin an emergency sourcing process for substitute parts, alternate carriers, or even new suppliers in time to at least prevent part of the forthcoming disruption, and pull all the data in from the relevant systems that you need to start that process.

It’s the only way to create a supply chain aware integrated sourcing strategy, which changes the static old world view of:

  • Identify the core product lines for the next 5 years, create the regional and local supply chain hubs, (blanket) contract the major carriers, and set up the infrastructure
  • Finalize the products for the next 1 to 3 years, do the sourcing events, cut the contracts, define the normal delivery schedules and routes, set up the (auto) reorders, and set the supply chain in motion.
  • Make minor changes based on localized demand variations, localized/temporary disruptions, or unexpected situations, get things back on track, and continue with the production and sourcing plan as normal

This worldview is unable to cope in today’s global economy that is plagued by pandemics, wars, major shipping route interruptions due to severe weather (Panamanian droughts) and rebels/terrorists (Houthis in the Red Sea), border closings and trade wars. While it more-or-less worked fine for the relatively stable global “free” trade of the first two decades of the century (except for when natural disasters occured), it’s clear the model no longer works and needs to be replaced by a better one.

More specifically, an integrated multi-level model that can react and re-run continuously as a result of signal changes to help the organization react and adapt to changes while still keeping long term and mid-term goals in mind with every decision. An ideal model that is spelled out in Part 4 of our direct sourcing series with Bob Ferrari of Supply Chain Matters on how integration is key.

For Proper Direct Sourcing, Different Organizational Thinking is Required

In our last post we noted that standard sourcing solutions don’t work for direct and referred you to our seven part series with Bob Ferrari of Supply Chain Matters at these links:

And we noted the reason was that direct sourcing doesn’t work isolated from supply chain. Fortunately, direct sourcing and supply chain planning can work together, but only if we

Think Different

This is the only way we are going to realize business and operational planning alignment from source to supply. Right now, it takes too much time across the various strategic, tactical, and operational decision making processes in the gathering, assimilation or transcribing of the most up to date line-of business, functional or operational focused data and information into spreadsheets and antiquated tools to support forecasting, sourcing, supply chain, and logistics systems.

This is primarily due to the fact that not only are each of these processes for different timeframes but they are typically conducted using different business processes. Long term strategic planning is typically conducted using IBP methodologies, mid-term tactical planning is typically conducted using S&OP methodologies, and short-term planning is typically conducted using exception planning, materials replenishment planning, logistics re-routing, etc.

Each plan requires information on the connecting layers in order to make a decision. IBP requires knowledge of what S&OP can do and the best historical results from S&OP to come up with the plans most fit for execution. S&OP requires knowledge of the overriding IBP goals as well as the operational systems used for day to day procurement, inventory replenishment and management, logistics and trade management, and production. Since most of these systems don’t talk, it’s a lot of manual data collection, processing, and pushing up and down the levels and the chain.

These processes need to be connected in integrated planning loops that span:

  • Plan, Source and Procure
  • Plan and Analyze
  • Plan and Produce
  • Execute and Fulfill

Moreover, these planning process frameworks need to be enabled by more effective data management, data harmonization, and analytics that enables these loops to constantly be executed and re-executed as needed to ensure each level of planning and each step of the process has the data it needs to suggest the right answer for the human to make the right decision.

Finally, this will only happen if organizational employees think different and adopt new processes, frameworks, data models, and strategies to integrated planning from source to supply. For some insights in to how this might happen, see part three of our joint series on how today’s Organizational Thinking is Wrong.

Standard Sourcing Solutions Don’t Work For Direct

the doctor recently teamed up with the Supply Chain Master Bob Ferrari over on Supply Chain Matters to bring you an initial seven part series on why Standard Sourcing Technology Solutions Don’t Work for Direct, which you can find at these links:

If you read Parts I and Part II in detail, which you most definitely should because we’re only going to summarize a few highlights here, we detail some of the big reasons they don’t work, besides the fact that most were designed for indirect and can’t even do the basics of direct sourcing. The reasons we put forward included:

  • Direct Material Sourcing is Hard
    • substitution (like satisfaction) is not guaranteed
    • substitution is always conditional when available
    • demand is not easily aggregated
    • delivery time guarantees are often significantly more important
  • Sourcing Platforms Don’t Do Direct Well (as most were designed for indirect)
  • Most Sourcing Platforms Don’t Support Bill of Materials
  • Most Sourcing Platforms Don’t Support Optimization

Then we dove into why direct solutions don’t work either:

  • It’s Not Just Landed Cost, It’s Total Cost of Acquisition
  • It’s Not Just Cost, It’s Supply Assurance
  • It’s Not Just Supply Network Assurance, It’s Timing

That’s just the baseline sourcing side of the equation. We still haven’t talked about the supply assurance side:

  • They Aren’t Designed for Multi-Stage NPD/NPI Sourcing and Quality Assessments
  • They Aren’t Designed to Capture Network Performance and Carrier Risk
  • They Aren’t Designed to Capture and Assess External Risks

That last point is key. If you’re not considering the geopolitics of where you are sourcing from and where you are sourcing to, and how those might change in the near future, you could be in for quite a shock, as many of you in the USA found out this year. If you had been paying attention to the election, noted how much a certain Tech Bro donated to a certain campaign, and compared that number to past campaign contributions, you would have known the election, which appeared neck and neck, was being bought and paid for, which party was going to win, and who was going to be President.

If you did your research, analyze everything he said publicly in the decades leading up to his first campaign for political office, look at what he actually did in his first term, and read Project 2025, you would have known something was coming on the trade front, especially where certain countries were concerned. And you would have known that what was coming was not going to be good for your business if you were sourcing from China.

But it’s not just the “to” destination you have to worry about, especially if the only thing increasing is cost. It’s also the “from” destination, which could be cut off entirely by a new regime that imposes sanctions or embargoes, or could undergo a rapid economic decline due to bad government decisions, external third party sanctions and embargoes, or global shifts in trade. A great discussion of this can be found in Koray Köse’s recent LinkedIn post on on Poland’s Economy: Reslient Amid Political Storms and how it faces a test under it’s new President — and how, should it fail that test, supply chain leaders need to be prepared. It’s the perfect example of why supply chain considerations need to be pulled back to sourcing, because there’s no way an average sourcing professional today would consider any of this when evaluating suppliers for a direct sourcing project.

When Someone Says “Real AI”, Ask For Details!

We shouldn’t have to remind you, but since too many people are falling for, and buying into, the hype and selecting tech that does not, and can not, ever,work, we are going to remind you yet again.

Computers do NOT think!

To think is to direct one’s mind … where one is an intelligent being, not a dumb box. Computers thunk … they compute using algorithms (which are hopefully advanced and encapsulate expert guidance and knowledge, but that is far from guaranteed).

Computers do NOT learn.

Appropriately selected and implemented probabilistic / statistical / machine learning algorithms will improve their performance over time as more data becomes available, but they do not learn. Learn is to acquire knowledge (or skill), and by definition, knowledge can only be acquired by an intelligent being.

Computer Programs Can Adapt …

but there’s no guarantee the adaption is going to improve their performance under your definition, or even maintain their performance. Their performance could actually decrease over time.

What is critically important is that there are two primary types of algorithms that can be used to create an AI application:

Deterministic and Probabilistic

A deterministic algorithm is one that, by definition, given a particular input will, no matter what, always produce the same output, with the underlying machine always passing through the same sequence of states. As long as you don’t screw up the input, or the retrieval of the output, (and, of course, the hardware doesn’t fail), it is 100% reliable.

A probabilistic algorithm, in comparison, is an algorithm that incorporates randomness or unpredictability into its execution, and may or may not produce the same output given successive iterations of the same input. Nor is there even any guarantee that the algorithm will produce a correct, or even an acceptable, input a given percentage of the time. Well designed, these algorithms may allow for consistently faster computation, better identification of edge cases, or even a lower chance of error, on average, for a certain class of inputs (but with the caveat that other classes of inputs may suffer a higher error rate).

Deterministic algorithms can be relied on to execute certain tasks and functions autonomously with no oversight and no worry. Probabilistic cannot. In other words, you cannot assign a probabilistic algorithm a task for autonomous computation unless you can live with the worst possible outcome of the algorithm getting it wrong. And this is what Gen-AI, and most of today’s “AI” tech, is based on.

This is the critical problem with today’s AI-tech and AI-Hype. Especially when a probabilistic system can, by definition, use any method it likes to determine a probability (which may or may not be at all appropriate, since a model is only valid if it accurately captures the “population” dynamics) and may, or may not, be accurate. For some of these situations, it will be the case that neither the company nor the provider of the system will have enough historical data (market situation and outcome) to even attempt to make a reasonable prediction, and there definitely won’t be enough data to know the accuracy, because standard measures of model accuracy (like the Brier Score), tend to require a lot of data, especially if you have a situation where you need to accurately identify rare events as this could require 1,000 or more “data points” (which, in a typical market scenario, would require enough data to identify the market condition and then the unexpected change”).

(And this is exacerbated by the reality that, for many of these situations, one could likely employ more traditional “statistical techniques” like trend analysis, clustering, classical machine learning, etc. to solve much of the problem at hand.)

It’s important to remember that Gen-AI LLMs, which power most of the new (fake) agentic tech, are all probabilistic based (and designed in such a way that hallucinations are a core function that CAN NOT be eliminated), and much of it is complete and utter garbage for what it was designed for, and even worse for tasks it wasn’t defined for (like math and complex analyses). (Everyday we see a new example of complete and utter failure, often due to hallucinations, of this tech. For example, you can’t even get a list of real books out of it — as per a recent contribution to the Chicago Sun Times which which published its Summer Reading List of 15 books, of which only 5 of which actually exist. And then there are numerous examples of lazy lawyers getting raked over the coals by judges for using ChatGPT to do their homework and quoting fake cases!)

While we do need to augment purely deterministic tech with more adaptive tech that uses the best “statistical techniques” to more quickly adapt to situations, we need to spell out the techniques and restrict ourselves to what is now “classic machine learning” where the algorithms have been well researched and stress tested over decades (not modern Gen-AI powered agentic tech that has worse odds than your local casino). At least then we’ll have confidence and can enforce bounds on what the solution can actually do (to limit any potential damage).

Especially now that we finally have the computing power we need to effectively use tried-and-true “classic” ML/AI techniques that require large data stores and huge processing power for highly accurate predictions. The reality is that even though this tech has existed for at least 25 years, the computing power required made it totally impractical for all but the most critical situations. Twenty-five years ago, a large Strategic Sourcing Decision Optimization (SSDO) model would run all weekend. Today you can solve it in a few seconds on a large rack server (with 64 cores, GB of cache, and high-speed access to TB of storage). The fact that we finally have (near) real time capability means that this tech is not only finally usable in all situations, but finally effective.

[And if vendors actually hired real computer scientists, applied mathematicians, and engineers and built more of this tech, instead of script kiddies cobbling together LLMs they don’t understand, we would be a decade ahead of where we are today.]