By Norman Katz, President Of Katzscan Inc., www.katzscan.com
It is a confusing time in the face of business technology. With the introduction and promotion of artificial intelligence (AI), machine learning (ML), and blockchain, it can be difficult for an enterprise to know what technology is needed, and when, in the lifecycle of the technology and the company’s own pathway. But in the race to progress forward, sometimes companies leave behind unfinished projects that are the necessary foundation for achieving success with these more advanced technologies.
These advanced technologies – AI, ML, blockchain – are typically associated with companies that are embracing supply chain digitization. But digitization – replacing paper-based transactions with electronic data transactions – really started decades ago when companies integrated their Enterprise Resource Planning (ERP) systems with Electronic Data Interchange (EDI), and when the scanning of barcode labels replaced pen-and-paper recording.
So, my opinion is that I don’t really buy into this whole new “digitization” wave, unless yours is a company that is struggling with implementing and integrating these 1980s technologies which are still the foundational technologies that run global supply chains today.
(I am an expert in ERP, EDI, and barcode technologies, and the supply chain business operations that they impact, so if your company is having troubles, feel free to reach out to me for help.)
Before you even think about AI, ML, or blockchain, ask yourself how smooth-running your ERP, EDI, and barcode scanning systems are functioning. Are you internal customers (system users) working efficiently or is paper still being printed for data entry? Are wasteful, non-value-added emails being sent because dirty data has prevented someone from performing their task? Are your employees frustrated with the software because of the software, or because of how the data was poorly setup within the software? And were the employees properly trained on the software features at the outset?
AI – artificial intelligence – is effectively pattern recognition of data. If the data is dirty, the pattern becomes corrupted. Imagine trying to extract meaningful analytical reports from bad data, let alone now layering ML (machine learning) on top of AI. ML is effectively software that will attempt to take lessons from AI and learn and automate and make decisions with minimal – or none – human interaction.
So, if the underlying data is bad, what is the lesson to be taught to ML? And do we want to trust ML to make decisions and take actions upon a foundation of bad or even questionable data? I don’t think so.
If your company’s master data management is so bad that it is having internal disagreements and power struggles over who owns (or does not own) what, then this is likely spilling out into your supply chain relationships with your customers and suppliers/vendors (e.g., contract manufacturers and distributors). As such, your company is the cause of its own supply chain disruptions. Evidence of this would be in the constant corrections to EDI transactions. (There would also likely be a financial impact via the penalties assessed, known as “chargebacks”.)
In as much as EDI transactions should only be changed via EDI transactions, manual intervention is commonplace, as are the communications via email and telephone via supply chain partners to come to a resolution as to what to do once the problem has been identified. Just imagine not being able to make alterations via blockchain relationships, whereby the transactions are effectively one-and-done via the “immutability” of the blockchain ledger. Is your company’s data and decision-making this accurate?
(Blockchains would have to evolve to allow a transaction to be further updated even if not directly changed, and I don’t know that such an evolution has taken place yet. Nor do I believe that layering changes upon changes to an original transaction is an effective means of transacting business.)
I am absolutely not against any of these advanced technologies. Rather, they offer enterprises tremendous opportunities to advance their employees beyond the mundane work, and to help the company excel in executing beyond their competition. But there is a trap here that businesses need to be aware of.
I am reminded of an old funny saying that goes something like this: “To err is human. To really mess things up requires a computer.” Before you consider embarking upon advanced technology projects, make very certain that your foundational software systems – those like ERP, EDI, and barcode scanning – are performing at their peak and are exchanging clean and controlled data, not just between those systems, but also between your company and its external supply chain stakeholders.
The lesson here is that AI, ML, and blockchain systems cannot and do not stand alone. They are extensions of core business system, the Enterprise Resource Planning (ERP) system. If the ERP system, and the systems it touches, are not able to support advanced technologies, those systems should be the priorities, not the new technologies.
Because if you want to cause chaos and catastrophe quickly, go ahead and layer advanced technologies upon foundational systems that are poorly implemented, partially integrated, and running on dirty data.
About the Author: –
Norman Katz is the President of Katzscan Inc. (www.katzscan.com), a US-based consultancy celebrating its 25-year anniversary in January 2021. Katzscan specializes in improving supply chain performance, business operating effectiveness, strategic software applications, and information insights. Norman is a multiple book author and a worldwide speaker and writer.