Why Your ProcureTech Initiative Will Fail … Part 1

We’ve written many series on best practice tech identification, tech selection, and tech implementation in the hopes of inverting the odds from an 80%+ chance of failure to an 80%+ chance of success, but given that failures are literally still happening on a daily basis, it seems that most people can’t be bothered to read best practice advice so today we’re going to flip the script and tell you all the reasons you’re going to fail and then you hope you go back and read the best practice advice we’ve freely given you (in series such as Successful Vendor Selection – The Series).

1. You Don’t Understand Your True Needs

You’ve never done a full end-to-end process analysis on your organization, you don’t understand how inefficient your processes are, what processes you actually need, why you need them, and how much better you could be doing. You just know that the KPI metrics you are tracking are not on par with industry averages based on what your overpriced consultants are telling you, that your balance sheet isn’t as good as best in class, and that you need to do something … and that something is get a shiny new tech toy that the overpriced consultants will help you select by telling you who to invite to your RFP. (And you should know all the problems with this — they’ll only recommend the partners they have sycophant partnerships with, get referral and implementation fees from, and who will ensure that they remain your overpriced consultancy of choice.)

2. You Don’t Understand What You Already Have

Once you understand what the correct processes are, why, and where the automation points are, you need to revisit the systems you have to see where they can solve the problems. Chances are you have a number of suites, supply chain platforms, and ERPs with easy to implement plug-in modules that solve a lot of the problems you have without buying any new systems. And even if new systems might do it better, chances are the improvement won’t be worth the extra money, downtime, and change management — which will all cost you dearly. The reality is that if you can get an 80% solution today, with tools your people are already using, that’s much better than a potential 95% years in the future.

3. You Don’t Understand What Your Capabilities Actually Are

By this we don’t mean your process capabilities or technological capabilities, we mean your actual functional capabilities. Your domain knowledge, your ability to execute on that domain knowledge, and your natural efficiency. It’s pointless improving processes to apply more advanced techniques or employing modern technology to speed up processes when you’re not capable of managing those advanced processes or technology. If you employ processes and systems you’re not ready for, they won’t deliver any results while costing you millions of dollars in the system selection and implementation processes.

4. You Don’t Understand How Long It Will Take to Upgrade Your Capabilities

Even if you figure out you need to upgrade your skills and those of your team’s, even if you posses a fair degree of human intelligence, you don’t know how long it will take. It’s not just buying a knowledge dump from a consultancy or giving your team a 5-day crash course, because knowledge that is not applied is not retained. There’s a reason College and University courses give assignments and projects as well as exams — the more you apply, the more you retain. If the imparted knowledge is not applied, it will not be retained. Until your team can start applying, repetitively, the new knowledge in improved processes, they won’t retain it and they won’t advance. The best training will be a day or two a month over months, not a week. And that’s for stage 1. It will take years to get your team from average to mastery. We’ve known for decades that major transformation projects take 5 to 10 years, and that the average journey to best in class for the committed is 8 years. Technology doesn’t change that. The longer you choose to ignore this fact, the longer you will fail.

To be continued in Part 2.

The Procurement Knowledge Devolution

The internet was supposed to kick off the procurement knowledge revolution, allowing Procurement pros to quickly:

  • find out about best practices
  • get commodity and market pricing for products and services
  • build reasonable (should) cost models
  • find out about new suppliers, carriers, and consulting partners
  • research new products and services
  • hold truly global sourcing and procurement events in real time
  • effectively communicate with, manage, and develop suppliers
  • etc.

For twenty years, that’s what it was. Procurement departments who learned how to use the internet properly for research, deployed the right SaaS to support their processes, and identified the right data for their processes were successful in their endeavours.

But then came Gen-AI LLMs, chatbots were replaced with chat, j’ai pété‘s and clod‘s, and knowledge was replaced with whatever content the LLM generated. Maybe it was correct, maybe it was mostly correct, and maybe it was a 100% fabrication — a hallucination if you please. The problem with LLMs is that they are NOT intelligent. They are essentially super sophisticated multi-level cross-connected deep neural nets that go beyond classification to generation of responses built up from sub-responses built up from deep training on incredibly large data sets.

Therein lies all of the problems. It generates built on random probabilities. Those are dependent on what’s in the training data set, what questions were asked, what results are reinforced, and how it’s used. If the training data is bad and full of bad data and falsehoods, the chances of incorrect, and even dangerous, responses being generated are quite high. If the training was biased, the output is likely to be very biased. And if it’s not “trained to please”, the models are fundamentally designed to “learn to please”, so if computations that are a complete lie will increase utilization of, and faith in the model, that’s what will happen.

Moreover, every vendor is now believing the hype from the big LLM players, treating the technology as real Artificial Intelligence (when it should be called Artificial Idiocy), and trying to plug it in everywhere … promising that it will provide their clients with true natural language interfaces, agentic tech, and even BS AI Employees. Those who are adopting it are literally getting dumber by the day. Not only has the cognitive impairment, atrophy, and potential long-term decline from regular use been well documented, but the failure rate has been well documented as well with MIT and McKinsey studies demonstrating success rates of 5% and 6% successfully. Most pilots are being abandoned, sometimes before they even begin, because the tech isn’t even good enough to put into employee’s hands for the tasks the overpriced Big X consultancies claimed the LLMs would be perfect for.

Furthermore, even when organizations are smart enough to ignore Gen-AI, over-automating using the most advanced last-gen (A)RPA and AI technologies will still give Procurement teams a false sense of security and, due to their very low error rate, as time goes on, the team’s skills will go rusty and their ability to deal with true exceptions quickly disappear, especially as the old Pros (get forced to early) retire and the younglings have never dealt with exceptional situations.

The age of AI hype has ushered in a knowledge devolution faster than any age that has come before.

I hope the profession can survive it!

Cost Reduction … it Starts With Cost Increase

It used to be cost reduction, which was focussed on cost cutting, started with the one-trick pony of cost cutting by any means necessary, which typically took the form of e-Auctions, RFPs to new suppliers, and GPOs that could aggregate and leverage huge volumes — all tricks that are rearing their ugly heads again with the rapidly rising costs thanks to inflation, tariffs, and global instability.

They all work just fine in the short term, but they all come back to bite you in the backside in the long term. Here’s why:

  • e-Auctions: find savings by squeezing margins, and you can only take those out once, and once inflation comes back, costs go up
  • RFPs: designed just to find the absolute lowest price attracts suppliers who cut corners, underpay their staff, and offer no service while alienating your current, more trustworthy, suppliers
  • GPOs: can aggregate volumes and lower prices, but then you are dependent on them, and paying their markup … forever

None of these is the long term answer.

When we first started discussing cost reduction two decades ago, the key methods we focussed on were:

  • strategic supplier relationships and customer of choice: so that they put the effort into being your supplier of choice and finding their own ways to keep costs down (streamlined operations, better raw material sourcing, etc.)
  • supplier investment and development: if the supplier is smaller, or not as advanced, they’ll only do so much on their own, so your efforts to invest, improve, and guide them (through early payments, low-cost new line financing, etc.) could greatly lower your costs over a multi-year engagement
  • strategic sourcing decision optimization: where you did a multi-objective optimization that took all of the cost factors (unit, transportation, warranty, service, waste etc.) into account as well as risk (that could cause “savings” to evaporate over night) and quantitative assessments of other key factors

And those are all good techniques in (semi) normal times. But these are not (semi) normal times. These are almost unprecedented times. Between natural disasters, geo-political conflicts and wars, and terrorism, we are dealing with unprecedented simultaneous reductions and closures of major maritime shipping lanes (the Panama Canal, the Red Sea, the Strait of Hormuz), unstable (and rapidly escalating) fuel costs, regular supplier and carrier failures, unpredictable crop and raw material availability, etc. all at the same time. Old friends becoming foes, or at least frenemies; friend-shoring, near-shoring, and home-shoring finally gaining ground (despite being promoted and the right answer for decades); and supply chains being swapped whenever possible.

We’re in times where these techniques, while still good, can’t always address all of the situations. Plus, if you’re constantly adapting to what’s available now, versus focussing on what you should be building, you’ll be in a constant, unstable, state of affairs, caught off guard with every flux, and constantly on the brink of ruin.

You need to stop working sourcing event to sourcing event, procurement to procurement, and disruption to disruption and start working on transforming your supply chain to a more resilient long term supply chain. This will require identifying which safe countries and regions (likely to have long term geo-political and trade stability with your home and/or destination countries) you should be doing business with, where solid supply bases could be, and how you could construct a real supply chain from the source countries to the destination countries that don’t depend on unstable source points.

Then you have to engage the carriers, find partners to help you manage the export and import requirements and take advantage of FTZs (free trade zones), build or acquire intermediate warehouses and cross-docks, and be ready for trade with the local suppliers. Those will typically include multiple suppliers you are not currently working with, and they may need to upgrade their production lines, operations, services, etc. to serve you to your level of expectation. This will incur costs that your suppliers and partners will need to incur, which will need to be passed onto you. Which means, in the short-to-mid-term, your costs will increase. But if you design the right, stable, supply chain networks that you can use for years (or decades), develop the right suppliers, and maintain volumes, as operations improve, up-front costs get amortized, and economies of scale get optimized, costs will go down, and with long-term agreements, over multiple years, your company will see previously unrealized savings while your peers see their costs go through the roof.

So if you want to save money, you better be prepared to spend.

We Know You Should Have A Lot Of Global Trade Data …

As per our last post, if you want to design and run a relatively smooth running supply chain, you need a lot of data. As per our post, at a minimum:

  • product compositions
  • supplier locations
  • route details
  • alternate products, suppliers, and routes
  • sanctions
  • denied parties
  • tariffs, export and import
  • taxes and recoverables
  • carbon/GHG
  • commodity market and product cost data
  • currency conversions and trends
  • natural disaster risks
  • man-made disaster data
  • consumer and market sentiment data

Which leaves you with two major problems.

  1. where do you get it?
  2. how do you manage it?

There are two choices on where to get it:

  1. data consolidators and brokers
  2. government / public sector sources

As much as possible, you want to rely on option b), because, in this AI-HYPE filled world, data is now the most valuable commodity, the data brokers know it, and even when they are getting it from a government / public source (for free) and then processing it for your consumption, they are charging a premium for it. A subscription to even a fraction of the above data could cost you more than the annual SaaS subscriptions you are feeding it into (not counting your AI token costs which are going to continue to increase without bound if you are unnecessarily relying on AI for tasks you do not, and should not, be using Gen-AI LLMs for).

So you want to use the cheap/free government and public sources as much as possible. While this sounds easy enough, every single source will be in a different format, with different access requirements, different update frequencies, and different levels of completeness. We’re talking about everything from Excel files to real-time json requests, with multiple types of authorizations, access protocols, and transport protocols.

This bring us to the second issue — how do you manage it?

You basically need a DIY data orchestration platform. And the reality is, the majority of today’s orchestration providers, despite their grand claims, don’t really do that! Out-of-the-box, you can only integrate solutions they have already integrated with, and only import any data they have previously mapped. Plus, they are limited in intake to what they have designed for. Most of them aren’t even as powerful as last generation data mapping frameworks that allowed a data analyst to integrate all of the data from various sources into one common, central, database. (Now, it typically involved creating yet another data warehouse / lake / lakehouse.)

What you need is a next-gen supply chain orchestration platform that was built from the ground up to allow you to plug, play, and orchestrate data sources as well as workflows and applications using pre-defined mappings, Web 3.0 Markup, standard terminology, and statistical AI (with known confidence) that will auto-map as much as possible, minimize what can’t be auto-mapped, and generate additional data objects and tables for storage where your systems aren’t already handling that type of data.

A next generation system that also incorporates auto-mapping, auto-schema-extension, and auto-data-orchestration alongside workflow construction and third party system integration using APIs, MCP, and other integration technologies along with all of the common authorization protocols. A fully dynamic data platform that can serve as the core of a next generation SCP platform. One that is a level beyond the majority of platforms calling them orchestration platforms today,

And, finally, one that understands how to organize, federate, and classify data for true analysis.

But that’s another post!

How Much Global Trade Data Should You Have?

As per our recent series on supply chain stability, and the need for visibility and audits, you need a lot of supply chain data. On your products, suppliers, and routes. As well as their products, suppliers, and routes. Down to the source.

But is that all you need — just suppliers, products, and routes down to the source?

Well, as per our previous series, you also need detailed data on alternate locations and other product offerings from your suppliers, alternate suppliers and their alternate product offerings, alternate routes, and alternate carriers that service them.

But that’s not enough.

Product existence is not availability, and availability is not obtainability. You also need to know if the countries, parties, or connected parties (owners) are sanctioned and if they use restricted or banned substances in their products or operations. If so, you can’t do business with them … and spending hours, days, weeks or months evaluating them just to figure that out when you try to order the first shipment.

But even if there are no sanctions or official roadblocks, it still might not be possible to acquire the goods you want from certain suppliers due to export and import tariffs. That means you need to know any export tariffs and import tariffs that will be applied, in the export country, import country, and any intermediate countries (that may not have [usable] FTZs).

Also, since some goods and services will be acquired, change hands, or used in the source countries, you also have to be aware of all the country, state, and municipal taxes that you will need to pay, as well as any taxes you can potentially reclaim.

This means we’re up to:

  • product compositions
  • supplier locations
  • route details
  • alternate products, suppliers, and routes
  • sanctions
  • denied parties
  • tariffs, export and import
  • taxes and recoverables

Is that all?

Well, in some countries you have to be cognizant of carbon/GHG production, so you need to be able to track how much carbon is produced in the production of each unit and be able to report that when needed, as well as control that if the country the good is being produced in, or sold in, has a carbon quota or carbon tax.

And it sounds like we’re nearing completion, but not really.

We left out cost data. If you have to find alternate products or suppliers, you really want to understand your costs. And not just market costs, should costs. Which means you should be tracking, when possible, raw commodity costs, component and part costs that allow you to estimate production costs on top of input costs, and currency exchanges, that allow you to build reasonable should cost models.

And while this might be enough to make perfect sourcing plans, we all know plans rarely turn into reality. So having the data to build the perfect hypothetical supply chain is a great start, but only in theory. In practicality, you need to know how realistic the supply chain is.

This means you need relevant supply line related data as well. Transport times, and risk of spoilage and theft are critical. As is natural disaster risk data related to storms (hurricanes, tsunamis, [fire] tornados), the possibility of floods, wildfires, blizzards; and earthquakes (or volcanic eruptions) in areas where they are common. You need to track this data to judge the potential reliability of supply chains.

Now we’ve added:

  • carbon/GHG
  • commodity market and product cost data
  • currency conversions and trends
  • natural disaster risks

While you think this may be everything, we’re sad to say that while it’s close, it’s not.

It’s not just natural disasters that matter, it’s man-made disasters. Mines collapsing, factories burning down, and unexpected supplier and carrier bankruptcies can also bring down your supply chains. So you need real-time event data.

Furthermore, while it won’t bring down supply chains, it can greatly vary demand and being aware of consumer sentiment, including mass displeasure and (potential) boycotts. If you’re selling consumables or creating high-priced electronics with a limited shelf-life, you don’t want to overproduce. And if there are indications of a demand surge, you want to be prepared so you don’t have costly stock-outs that could result in unhappy customers.

That’s quite a lot of data. And that’s a minimum.