30 Years. Since January 1996. X12-EDI. ERP. Barcode. Author. Speaker. Writer.
30 Years. Since January 1996. X12-EDI. ERP. Barcode. Author. Speaker. Writer.

Not posted yet.
Not posted yet.
Not posted yet.
Not posted yet.
Not posted yet.
AI apologies only make matters worse.
On two separate projects, software vendors used AI to craft their apology statements after each caused disruptions to our projects. In both cases the use of AI to wordsmith their apologies was a public relations disaster for the software companies. The apologies were both rejected by my clients as disingenuous.
Both times, the AI apologies were instantly noticed by my clients, at one company by the controller and at the other company by the chief financial officer.
The financial executives called out the wording, noticing that the person who delivered the apology could not have thought up the wording that was used given the history of the communications with that person. Another giveaway was that the apology’s font was a mismatch to the font of the rest of the email. And in one apology, the controller called out that the AI’s “disclaimer” was copied-and-pasted in along with the response.
The controller vocally (in an email reply) called out the software vendor for their use of AI. The software vendor attempted to defend their use of AI for its “wide use across all industries” and “when writing emails, it ensures addressing all relevant points, save time on writing and reviewing emails while keeping a professional but also friendly tone. While it saves us time, it saves our customer on billable time on writing emails and communication.” We’re pretty sure that AI was used to craft that response too.
On the other project, the chief financial officer even knew which AI was used and essentially “called B.S. on everything” regarding the response as related to the larger topic being discussed.
AI is a revolutionary technology, but like any software it has its proper use. Soft skills still rule. If you need help writing something, you are better off finding your own voice rather than relying on copying-and-pasting what AI has to offer, because it will only come off as fake, and this will erode trust and can eventually end a business partnership. If you need to use AI to get ideas or a framework, that’s fine. But say what you need to say yourself so it is conveyed and received as genuine and sincere. You’ll preserve your business relationship better that way.

US cybersecurity spending: more buck, less bang.
The 2025 Cybersecurity Maturity Report – using data from 17 countries and 15 industries – reveals that despite an increase of 15% in cybersecurity spending, companies are less secure and at a higher risk of a cyberattack.
Companies are unable to fully assess their IT assets, keep their software up-to-date, and are unable to convince their employees to use strong passwords. The report noted that a shocking 75% of surveyed companies do not have full visibility to their IT assets.
Japan and Norway – with well-planned national cybersecurity strategies – fared better than countries like the US and UK who are without such national strategies. This proves that strategic planning can be more instrumental than a bigger budget in addressing cybersecurity.
According to a report by Verizon, third-party risk is a vulnerable spot for supply chains. In 2025, 30% of breaches were due to weaknesses in third-party systems, yet companies are still failing to address this supply chain gap.
And per the report, 50% of the companies surveyed do not have a disaster recovery plan in place.
Based on the statistics presented in the article “Why supply chain cybersecurity still falls short – and what leaders must do next” by APQC in the 2025 September/October issue of Supply Chain Management Review magazine: (a) the average number of calendar days to detect a cybersecurity incident is 186; (b) the average number of calendar days to respond to and recover from a cybersecurity incident is 45 days; (c) the average number of calendar days to notify customers of a breach is 30; (d) the average number of calendar days to apply security patches after they are released is 30. These numbers are each way too long and leave the breached company and the compromised data vulnerable to even greater risk exposure.
Hackers will find and expose the weakest link to gain access to a company’s systems. Whether by planting a virus or via social engineering, regardless of whether directly infecting a system or indirectly through a related third-party, criminals are smart and patient as they pursue their prize. Cybersecurity relies on investing in the right protective hardware and software, but people are most often the vulnerable access points.
Companies need to invest in constant training and education to keep employees sharp and focused on security.
Be vigilant. Be smart. And think before you click that link.
Small suppliers are wanted by big buyers
A report released by Supplier.io last year revealed that 96% of the more than 150 surveyed executives from companies with greater than $500M in annual revenue stated that they would like to integrate small suppliers into their supply chains. Unfortunately, these surveyed executives cite three challenges in accomplishing this objective.
Visibility to small suppliers, vetting small suppliers, and onboarding small suppliers were the primary roadblocks to engaging with more small suppliers.
The first challenge – visibility – is a tough one to overcome. Small suppliers can have few avenues for outreach with limited budgets for marketing. This is where I think that buying companies need to do a better job in reaching out to small suppliers. Memberships in local manufacturing associations and chambers of commerce can help, for example, but it takes both parties to actively participate. Small suppliers need to think creatively about how to reach out to large buyers, and large buyers need to be more open to expanding their own visibility to small suppliers.
Vetting small suppliers is a challenge and can be a burdensome or overly complicated process. There are various online platforms that claim to help streamline the process, but that doesn’t alleviate the need for suppliers to provide supporting documentation. Small suppliers may not be comfortable revealing detailed financial data, but they need to understand that large buyers have to be confident that they are going to be stable and reliable companies to do business with.
Onboarding small suppliers may be challenged by the fact that small suppliers run legacy software systems and business processes that are not capable of synchronizing or adapting to the technical and operational requirements of the prospective large customer. (I encounter this frequently with consumer product companies who are suddenly surprised by retail or grocery supply chain vendor compliance requirements.) Small suppliers should review the onboarding requirements beforehand, and also inquire about the technical and operational supplier requirements preferably before but certainly after onboarding to ensure that they understand what is going to be required of their company. The small supplier must weigh whether investments in software and business processes will yield the return on investment to gain the business customer, with the consideration that if one big customer can be acquired, why not others?
As noted in the survey, over 40% of companies reported a small supplier failure in the prior year. Small suppliers can improve confidence in their company by investing in better software, business processes, and targeted marketing. The payoffs can be substantial and help the company grow into new and larger customers. Just don’t take on more than you can deliver.
AI versus workforce development
An October 2025 report by the BSI (British Standards Institute) found that 40% of companies are using AI to reduce headcount and over 30% of companies are considering using AI before hiring persons to perform the same tasks. This analysis has led the BSI to the conclusion that AI automation “significantly overshadows” workforce development and the upskilling (training) of employees, new or existing.
The survey was conducted to a group of more than 850 business leaders.
The BSI noted that there was a difference in how small, medium, and large companies viewed AI and the relationship to the workforce.
While large companies were keen on using AI to reduce workforces, small and medium sized companies had little interest in doing so. The BSI survey found that only 30% of small and medium sized companies reduced junior roles in comparison to large companies.
The conclusion of the BSI survey is that small and medium companies may represent better opportunities for entry level jobs and foundational skills development. Consider that the Fortune 1000 is called that because it comprises 1,000 companies. But there are over 650,000 businesses in the US with between 20 and 499 employees according to a June 2025 SBA (Small Business Administration) analysis and over 5 million small businesses in the US with between one and 19 employees.
Yes, the companies in the Fortune 1000 have lots of employees, but there are lots of small and medium businesses representing lots of opportunities for employment.
While AI represents the opportunity to automate repetitive tasks and enhance analysis, the loss of entry level employees removes the foundational level workforce that should be trained and upskilled to move into more senior roles over time. What’s missing here is that these employees will have retained institutional knowledge along their journey with their employer.
Companies seem to be forgetting the value in that resource.
Entry level job seekers may not have the soft skills necessary for the job roles that they are applying for. This is probably true for jobs in large corporations and will probably be more noticeable and a greater factor when applying for jobs in small to medium businesses where a job seeker is more likely to interview with an executive or an owner. Updated resumes, improved LinkedIn profiles – including professional pictures – and brushing up on verbal and written communication skills will be critical in winning a job in a competitive marketplace where the competition is not just other job seekers but also against AI.
Acquiring a company? Check that data.
While helping a consumer product company conform with some supply chain compliance requirements, I noticed something that looked amiss with their item data, specifically the item data attributes associated with GTINs (Global Trade Identification Numbers).
GTINs are assigned by the manufacturer and provided to retailers and grocers, acting as a translation item identifier between the manufacturer’s item number and the retailer or grocer’s SKU (stock keeping unit). The most common GTIN is the UPC (Universal Product Code), the 12-digit barcode on consumer products. GTINs are also used to represent case (carton) packs. Master data management of item GTINs between the manufacturer, their 3PL or distributor, and the customer (retailer or grocer) is a key function in supply chains like retail and grocery.
Investigating further, and verifying my findings using the check digit calculator on the GS1US website, I discovered that the check digits on many of the GTINs in a subset of the item data were indeed invalid. When I presented this to my client, I was informed that the item data was acquired during an acquisition and simply merged into my client’s existing ERP system “as is”. No one had checked the validity of the item data including the GTIN attributes.
Invalid GTINs aren’t just an internal issue; they are an external problem as well. GTINs are typically included by customers in the line-item data on the X12-EDI purchase orders (EDI850) and sent back in the EDI856 ASN (Advance Ship Notice) and EDI810 invoice. Surprisingly, my client’s customers – retailers and grocers – never checked to validate whether the GTINs they were receiving were valid.
Barcode check digits are used by scanners to validate the contents of the barcode. If a check digit is invalid, the barcode may not be readable by a scanner.
The important lesson here is that data must be validated before being merged from another company such as during an acquisition. You don’t know how well – or how poorly – the acquiring company did in managing their data and ensuring that it met industry or regulatory standards. Don’t just accept acquired data on face value: run audit programs to ensure that the data your company is accepting meets your company’s data standards before you merge the acquired data into your company’s ERP or other business system. And if your company doesn’t have data standards, develop them.
Analytics before AI
During the vendor selection phase of an ERP (Enterprise Resource Planning) software project last year, one of the software vendors was touting the AI (Artificial Intelligence) capabilities of the ERP system they were representing. My client commented that they were interested in exploring these features. I jumped in and immediately shut down this conversation, after which my client agreed that it was too far a reach.
My client was replacing a 35-year old accounting system that had essentially been so heavily modified over the decades that it could no longer perform accurate mathematics. Purchase orders and invoices were created in Excel templates. The monthly financials were double-entered into Excel spreadsheets because the reports from the accounting system had failed to be accurate for years. The only printer the Unix-based system could support – a dot matrix printer – was visible behind the Operations Director on video meetings.
As I explained to my client and reinforced in software vendor meetings, the initial objective of migrating to a modern ERP system for this company was – among other basics – to simply get accurate analysis reports direct from the software. Forget about AI, we just needed analytics that my client could rely on and that didn’t require rekeying numbers into Excel worksheets.
I’ve written a lot about AI in my editorial articles for Supply Chain Management Review. They are cautionary tales about what AI cannot do, or about how not to implement AI. Here is a company that has never had good system-generated analysis reports … let’s start there before we jump to AI.
And without the need for much consideration, my client completely agreed with me. AI wasn’t mentioned again throughout the project.
A Stibo Systems whitepaper released in 2025 reveals startling statistics about the sad state of customer data. Surveying 500 business leaders in the US across various industries, the results reveal that 69% do not fully trust their customer data with 48% finding it incomplete. Over 80% of business leaders are not comfortable making sales forecast, pricing, or inventory decisions based on currently available customer data. The whitepaper is worthwhile reading for its statistics on how bad customer data is and the negative impact it has on execution.
If your company is considering implementing AI, think about whether you first have good analytics. If not, you probably don’t have good data. Without good data and good analytics reporting, how will you be able to judge if the AI output is in line with expectations or parameters? What will your check-and-balance be as you implement the AI?
Recklessly leaping into AI is the wrong approach. Cautiously stepping into AI is the better way. Data is the foundation.
30 Years.
This month is not only the beginning of a new year; it is also the 30th anniversary of Katzscan Inc. Tempus fugit. I honestly don’t know where the time went. As I reflect on what I’ve accomplished, it’s a lot.
Three published books, the first two available in hundreds of university libraries worldwide. Last year I had the pleasure of being interviewed by a master’s degree student in Hamburg, Germany who read and incorporated one of the themes of my first book – Detecting and Reducing Supply Chain Fraud – as a core part of his thesis. His master’s thesis scored an 88 out of 100 after which he graduated with his master’s degree in Global Logistics & Supply Chain Management from Kuehne Logistics University.
Two peer-reviewed papers for Henry Stewart Publications’ Journal of Supply Chain Management, Logistics, & Procurement. Both papers passed the peer-review committee on the first submission.
I’ve been a monthly editorial writer for Supply Chain Management Review – one of the leading supply chain publications – since 2023. My article series on The Perfect Order was published last year by SCMR. And a nice highlight last year was when Costco’s Connection magazine published my article “Don’t get foiled, what business leaders can learn from the sport of fencing” based on my third book Attack, Parry, Riposte: A Fencer’s Guide To Better Business Execution in their July 2025 issue for the For Your Business article.
(You can find links to all of my articles – including some PDF downloads – on the Katzscan website on the Articles & Audio page which is under the Media menu link.)
Over 60 US and international presentations. Over 60 US and international articles.
All of the above while helping leading companies with their software (selection, implementation, utilization), supply chain (compliance, financial penalty chargeback reduction, relationship improvements), data (analysis, mapping), and business operations (increased efficiencies, documentation).
It’s been quite the 30 years, much more than I could have imagined when I started out.
I hope you’ll take the opportunity to look around the Katzscan website and see for yourself what I’ve been up to these past three decades.
And if the past is prolog, I’m not done yet, as there is much more to look forward to achieving and many more companies for me to help.
In appreciation to all of those who helped me get to where I am today.
"Things seen differently."