28 Years. Since January 1996. EDI. Electronic Data Interchange. ERP. Enterprise Resource Planning. Supply Chain Vendor Compliance.
28 Years. Since January 1996. EDI. Electronic Data Interchange. ERP. Enterprise Resource Planning. Supply Chain Vendor Compliance.
Not published yet.
Not published yet.
Not published yet.
Not published yet.
Whether your AI is Discriminative or Generative, it still has to be Collaborative
Generative AI such as ChatGPT® - the GPT meaning “Generative Pre-Trained Transformer” – is of a type of AI that is used to create something, e.g., text-to-text, text-to-image, text-to-video. Generative AI is all the rage and gets a lot of notice in the media, for good and bad reasons. Yes, it is pretty nifty how one can create by prompting software, but it is also more than a bit disconcerting how easily fakes can be created.
Discriminative AI seeks to distinguish between different states or statuses, e.g., something that is good versus something that is bad. This type of AI can be used in manufacturing to determine whether what is being made was produced to required tolerances or minimum specifications. Items on the manufacturing line or going through quality assurance that do not pass inspection criteria can be automatically routed to a quarantine area. Discriminative AI can also be used to flag transactions that are out-of-bounds from what would be considered normal transactions. This type of AI is therefore useful and helpful for transactional fraud detection. (Remember: AI tends to be objective, so a person with a subjective perspective should assess the AI’s callout to conclusively ascertain if illicit activity is going on.)
Regardless of which type(s) of AI your organization implements, remember that every type of AI needs to be collaborative. The AI you implement needs to work along with:
a. Your business data – Generative AI creates new content based on what it knows about existing content and what it learns. Discriminative AI knows the difference, e.g., what’s good and bad, based on how you train it.
b. Your business software – If the AI needs to interact with business systems like ERP (Enterprise Resource Planning) to inquire on transactions, the transactions need to be reviewed and ensured to be usable by the AI.
c. Your business processes – The use of AI needs to be an integrated part of business processes.
d. Your employees – Determine how your employees will interact with the AI and make it a part of their daily job experience.
e. Your external stakeholders (customers, suppliers, vendors) – Determine how your customers, suppliers, and vendors will interact with the AI. Just because there may not be a direct interaction does not mean that there may not be any interaction. For example, are you using discriminative AI for quality control of inbound goods received from suppliers/vendors?
f. Your policies – The use of AI should be disclosed in your ethics statements. Your customers and suppliers/vendors should be aware of your use of AI just like your data collection practices. You don’t have to reveal company secrets, but you should advise – clearly and proactively – when and how you organization is using AI.
Like so many other technologies, AI has the potential to offer great promise in helping organizations get more done more quickly. But it has just as great a potential to be abused, misused, and too-trusted whereas it becomes a liability instead of an asset. AI is in its early development, so if you can find a good use for it, do so, while proceeding with the right caution and expectations.
Customer Loyalty versus Customer Longevity
Does it just seem like you’ve been with some companies for such a long time even though the relationship is a rocky one, and you don’t know why? It may be an insurance or services or technology company – something other than a monopoly (e.g., a utility provider) – with whom you just keep renewing year after year even though you’d like to change. Despite being unhappy with their customer support, disappointed with their product or services, upset with their website, or frustrated with their offerings, you just stick with them, and you’re not really certain why?
A friend of mine articulated it best recently when I was complaining about this situation by reminding me of an age-old saying, which I’m paraphrasing here: Sometimes the devil you know is better than the devil you don’t.
I think companies make the mistake that customer longevity equates to customer loyalty, but in my opinion, those two concepts are very distinct from one another. With the devils that I know, I would not likely be interested in any of the other products or services that those companies are trying to sell to me. If these companies cannot successfully, or even passably, complete their core missions that I am depending upon, what makes them think that I as a customer am going to trust them with anything else that they are marketing to me?
Nor why would I spend time to complete satisfaction surveys if I am already angry with a company? It might help to blow off some steam, but after so many survey requests and commentaries where nothing has changed over the years, why would I continue to waste my time with these surveys, many of which miss the mark or are overly complicated, some delving into questioning me on my personality traits? I just delete those emails.
Senior executives may get the wrong impression that no news is good news, and that customer longevity implies customer loyalty, but I contend that those are more likely false signals, and that a better detailed analysis should be performed to really ascertain the difference between customer longevity and customer loyalty.
Typically, companies have the resources to do better in what they should be doing, but flatly refuse to do so. Loyal customers engage more, and that translates to higher profits such as royalty commissions and affiliate fees from customers who use more services and buy more products. But disengaged customers are more likely disinterested in what the devils-they-know companies are spending marketing money to peddle.
Silence is not always golden; instead, it may be the indication that something is wrong. If you’re not hearing more from your customers – especially the ones you’ve had for a long time – that may not necessarily be an indication that everything is running smoothly. It might also be a warning that your customers just view you as the devil they know, and are waiting around for an angel to take your company’s place.
Don’t just perform, execute.
The terms “performance” and “execution” have a tendency to be used interchangeably, but I view them as being different, with performance as something that is a part of overall execution.
To me, performance is something measurable, represented by metrics, and defined by operational or technical activities. Did something happen early, on-time, or late? Was something (e.g., an order or an order line quantity) complete versus incomplete? Was data correct or incorrect? Did a barcode label scan or not scan? Performance can be quantified and analyzed.
Execution – to me – goes beyond performance, because execution cannot always be directly analyzed and assessed for its impact. Somewhat akin perhaps to goodwill in accounting in that it includes the intangible, execution represents not just performance but also the aspects like customer touchpoints and check-and-balance reporting: activities that are important in keeping good relations or maintaining good controls, even if the exact value or cost savings cannot be directly determined.
Companies can use reporting from both their ERP (Enterprise Resource Planning) and EDI (Electronic Data Interchange) software systems to inform of orders, shipments, and invoices. Together, reports from these two key and inter-related (even if not integrated) business systems can act as a check-and-balance with each other, informing of not just activities but also of errors, updates (such as purchase order changes), and critical information on a timely basis. Reacting as soon as possible rather than too late can have a direct and positive impact on business customer relations and keeping chargebacks (financial penalties for supply chain non‑compliance) as low as possible.
If shipping direct to consumers, sending emails or text messages informing of the orders status along the order fulfillment flow (pick, pack, ship, deliver) is a nice way to keep the consumer informed without over-communicating. This proactive communication alleviates the consumer from having to reach out to your company to inquire or wonder about where the order is and having to check for themselves.
Performance metrics and key performance indicators are important to help a company understand where problems may be occurring. Execution may be difficult to measure but does help to reduce excessive operational costs and keep customers happy, and coming back.
This doesn’t necessarily take a costly investment … some of the tools like reporting already come with the software systems company’s own, it just takes an understanding of how to utilize them strategically. Whether on-premise or cloud-based systems, start with the reports and data extracts you already have or can already create.
In a commoditized world, execution – not just performance – is the true competitive edge.
The whisper game is fun for kids, not for business.
Perhaps with so much technology being introduced at a younger and younger age, not too many people remember playing the whisper game when they were kids. It’s a fun game because it is simple to play and the outcomes can really be outrageous. Kids both hear and say the darnedest things.
The whisper game is played by taking a group of people (children, for the most part) and placing them in a line or in a circle. To start, a selected person thinks up (or is provided) a saying to whisper into the ear of the person next to them. It is whispered only once. That person hears what they hear, and whispers – only once – what they heard to the person next to them, and so on and so on, until the last person speaks out loud what they (think they) were whispered.
The comparison between the end result and the original phrase is often quite different, humorously so. Again, it’s really funny when this is done with kids.
In business, the whisper game doesn’t work very well. The results are often counter-productive and can lead to bad decisions.
The whisper game in business extends beyond verbal and written communications between people; it also includes data transformations between software systems as when two systems are integrated together, such as ERP (Enterprise Resource Planning) and EDI (Electronic Data Interchange).
Whether the data is flowing into (e.g., orders) or out of (e.g., shipments, invoices) your ERP system, make certain that you map directly to each customer’s individual business transaction requirement. The problem with converting inbound and outbound files (repeatedly) before the mapping process is that this is akin to playing the whisper game: something invariably will get lost in the process. In the case of the whisper game, it is the words that make up the original phrase. In business, it is the risk of losing or corrupting some of the original data that could have been really important to the whole of the transaction.
The more transformations or conversions along a file’s path, the more risk of problems and less reliability to the content of the file. The only screaming at the end of the whisper game should be by kids who are laughing at the funny result. Make certain that you reduce incidents of frustration and data integrity degradation by limiting the number of transformations or conversions you send a file through. Keep the message clean so that your transactions are heard loud and clear by the software system receiving them, whether it’s yours or belonging to your supply chain partners.
It’s still really cloudy out there.
According to analyst firm HFS Research (www.hfsresearch.com) who collaborated with EY to survey 500 senior executives from Global 2000 firms, of the two-thirds who made investments in cloud systems, less than one-third were satisfied with the return on those investments. Ouch.
Apparently, according to EY, “half of cloud-native transformations are abject failures”. Double-ouch … and why?
Based on the HFS Research: “The biggest challenge in capturing value from investments in CNT (cloud-native technologies) lies predominantly in aligning technology and business priorities to develop a coherent strategy.” (https://www.hfsresearch.com/research/only-a-third-of-enterprises-are-realizing-their-cloud-ambitions/)
Isn’t this what technology projects are supposed to do, whether cloud-native or not? Isn’t this what proper business analysis does: align the technology and the business needs? And isn’t this what good master data management is being partially about: the strategic – and sometimes creative – setup of data within a business system, e.g., to mitigate if not eliminate modifications and enhance functionality?
It seems that the lessons of the past have been completely forgotten just because we have systems in the cloud rather than systems on premise.
The conclusion of HFS Research survey seems to almost be that this is a technology issue, but why was that ever the core focus in the first place? We need to start with the premise that technology is a tool to get a business task done, not the other way around. If technology cannot solve a business problem, then the technology has no use for the business. Instead of leading with the technology focus, the conversation should prioritize the operational purpose. CIOs are under pressure from their CFOs and CEOs to justify the costs – and cost overruns – of these projects. Tech for the sake of tech is the wrong approach. And I have long believed that the role of the CIO is misaligned and co-mingled: let CIOs focus on the role of information and software and elevate the CTO to be equal in their oversight and voice of the technology platforms, including business software where operational fit is equal among different vendors and technical platform is critical.
Forgotten history is doomed to be repeated. It seems that the basics of software system implementation are missing from these projects, because the reasons for failure – cost overruns, skills gaps, operationally inefficiencies, business alignment – could have been pulled from the pages of software system failures from the 1990s, 2000s, or 2010s.
Security over the internet is different than on-premise systems of the past, granted. (And this is exactly where the CTO should be involved!) But, the foundations of what makes a business system implementation succeed or fail haven’t really changed: it all begins with what the business needs in the current state and in the future state. It takes an assessment of the systems in the marketplace to select the best fit. It takes creativity to bridge data and operational gaps to mitigate if not eliminate modifications. It takes business stakeholder engagement from the beginning through implementation. It takes clear communication, documentation, testing, and training. It requires an assessment of the potential solution’s technical architecture. And it requires an understanding that this is a business – not technology – application and project first and foremost.
Our hate-love relationship with outliers.
According to the obituary provided by The Washington Post, transplant surgeon Roy Calne passed away on January 6 earlier this year at the age of 93. Were it not for the pioneering efforts of British surgeon Dr. Calne, organ transplants that are relatively common and successful today would realistically be unheard of and not attempted due to the lack of innovation and success that Dr. Calne pioneered and persevered to push through.
Roy Calne was the son of an automobile engineer father and a homemaker mother. His inquisitiveness for wanting to understand how to take things apart and put them together comes from tinkering with cars in his father’s workshop. He also didn’t like being told that something wasn’t possible to do. Thus, equating replacing human organs with swapping car parts, he asked the question “why not”?
His initial experiments in the 1950s on animals were not successful, and he was threatened by animal rights activists. But Dr. Calne persisted, collaborating with another medical professional in the United States, Dr. Thomas Starzl, who similarly believed in the theory of organ transplantation and who performed the world’s first successful liver transplant. Dr. Calne performed the first successful liver transplant in Europe.
Calne led in the use of a new drug that was developed by a Swiss pharmaceutical company – Sandoz – called cyclosporine for suppressing the immune system to increase transplant success. (Immune system rejection was one of two major stumbling blocks Calne initially faced, the other being the surgical technique of how to replace one organ with another.) Starzl discovered FK‑506, another immune system suppressant that also contributed to surgery success.
Roy Calne enrolled in Guy’s Medical School in London at age 16. At age 21, he was hired to teach anatomy at the University of Oxford. At Oxford, he was attending a lecture by future Nobel laureate Peter Medawar on skin graft results between mice. During the lecture, Medawar stated that the experiment revealed the immune system could be manipulated, but that there was “no clinical application whatsoever” in doing so. Hmmm … my goodness … how many times have we heard such prodigious prognostications that turned out to be completely inaccurate? Good thing that Calne didn’t heed that statement.
Dr. Calne was knighted in 1986 for his breakthroughs in medical science. Calne and Starzl received the Lasker Award (medical science’s highest award after the Nobel Prize) in 2012.
Many people do a great job of dismissing – and sometimes vocally disapproving of – people with different ideas and different talents because they don’t fit into pre-defined molds or expectations. And yet, when these outcast outliers create something unique or solve difficult problems, they are heralded as heroes by the same people who initially discarded them. Here, an automobile tinkerer applied the same rational to fixing a person that he did to fixing a car, and he changed the world.
People with range – I strongly recommend the book Range by David Epstein – will give you more talent for your human capital investment. The resumes and online profiles of people with range will be easy to dismiss, but don’t. You’ll likely need to take a bit more time to read through and digest the careers and skills and experiences and qualifications of people with range, but in the end, it will be well worth it when it comes to finding a person who can creatively resolve the problems that are holding back your company’s progress. People with range say “why not?” when others only say “why?”.
In 2022, there were 157,494 organ transplants performed worldwide according to www.transplant-observatory.org. The lives extended and saved are really all thanks to the pioneering efforts and persistent pursuits of Dr. Roy Calne and Dr. Thomas Starzl.
The other economic theory I learned in university.
Inasmuch as the value of a university degree is recently being called into question, and I don’t dispute the reasoning, back when I attended it was pretty much a given that if one wanted to progress forward in a career, one should achieve their higher education certification.
My coursework marginally prepared me for the real world. Yes, I learned COBOL, but not from a practical programming perspective as was readily pointed out on my first corporate job. I suffered through both the degree-major weed-out accounting and finance courses, and struggled through two engineering calculus classes. I was never really sure why a computer science major through the business college was required to take those courses which were changed for business majors after I graduated. The required computational matrix mathematics class – I may never forget “random variables” despite not knowing what they are – caused one student to get sick just being handed an exam.
Through four economics classes I remember two economic theories, one of which is supply‑and‑demand. This is ironic I suppose since I ended up in a supply chain career of sorts.
But the economic theory that really stuck with me and made more of an impact through the years – personally and professionally – is the theory of diminishing returns to a factor.
Basically, the theory of diminishing returns can be summarized as stating that more of something has less value at some point, even when the thing you have more of has value.
For example:
· When you are sick, taking all of your medicine at once does no good, and in fact, it can be very harmful.
· Using too much laundry detergent can overflow the washing machine.
· Eating the right amount of your favorite food daily is fine, but if someone gave you more and more of your favorite food to eat daily, that extra food would be a burden, e.g., you would have to give up some other food to eat it, it would be too much and make you sick, it would go to waste, you would expend time and effort to give it away.
· No matter how fast your vehicle can zoom from zero to 60, most highway speed limits are 65 miles-per-hour. And with increasingly excessive traffic these days, you’re not likely to be able to drive solo on roads like in the commercials that convinced you to buy the vehicle you’re driving. So, how much horsepower under the hood is really required?
· “Love is like oxygen, you get too much you get too high, not enough and you’re gonna die.” (Thank you Sweet, for these sweet song lyrics.)
What if you had $1B (billion) dollars? That would be great! You could purchase 100 $1M homes for you closest 100 friends. (Um … now you have to figure out if you even have 100 friends for whom you would purchase a $1M home.) And to help them pay ongoing expenses like property taxes and insurance, you will set them each up with a $1M high-interest bank account. This generous outlay consumes just 20% of your fortune, leaving you $800M. Now what do you spend your fortune on?
Before your company madly rushes to purchase and implement advanced software like Blockchain, BI (Business Intelligence), ML (Machine Learning), and AI (Artificial Intelligence) to solve its woes, ask yourself what is really the root cause of your company’s problems? Because none of these software applications will correct bad data, a poorly selected or implemented ERP (Enterprise Resource Planning) system, or bad (or a lack of) business operational procedures. More has less value when the foundation is faulty, has failures, is limited, or is non-existent.
Advanced technologies, when they are ready, when your company is ready, have their place, when the foundational data and the independent and the connected systems have integrity and validity, and when the business processes have their gaps closed and are reliable. Until then, more just has less value, and will likely make things worse because what you think has value is only likely hiding the real problems, or actually making them worse.
Taxing ethics.
I owe so very much to my “Aunt Shirley” who turned the chronometer and hit her milestone 95th year last year. Shirley did my first (and all subsequent until her retirement) tax returns, typed my first resume, and thoughtfully created the name of my company. She was and still is a great supporter and inspirer. Shirley always stood on the side of ethics and integrity, even at the cost of revenue and risk of losing clients who would have her manipulate their tax returns. Retired and having passed her accounting firm to the next generation in her family, we continue to keep in touch as I am thankful that my personal and business accounting and tax needs are in good and trusted hands.
I love you, Aunt Shirley.
According to the November-December 2023 issue of Consumer Reports magazine, a July 2023 Congressional report discovered that three popular online tax return providers had been sharing the information of as many as 20 million customers with tech companies for years.
The information provided by H&R Block, TaxAct, and TaxSlayer to tech firms like Google and Meta (Facebook) and others included incomes, refund amounts, and names of dependent children of taxpayers. Scrambled information included names, emails, phone numbers, and gender. According to experts (per the article) the scrambled information could easily be unscrambled.
The Congressional report determined that both the tax companies providing the data and the tech companies receiving the data likely violated taxpayer rights and also may have violated privacy laws.
Likely … may have … but did they? Apparently, what was done is inconclusively illegal at the time of this writing. But was it ethical?
Just because something is legal to do that does not make it ethical to do. And just because there is profit to be gained, that doesn’t justify the penalty of the fine that might be paid if caught, which these companies were, even if they can afford it financially and shrug it off from a brand perspective.
Ethics means, quite simply, doing the right thing, like full disclosure and relative distance. If you’re not certain, speak up, and let transparency rule. If you think you’re too close to a decision to be made, like whether a friend or relative should get a job position or a contract, back out of the decision-making process. And, like old food in the refrigerator, if you’re not certain, you probably shouldn’t … like sharing (or more likely selling) what should be confidential data.
Apparently, companies continue to believe that acting ethically means that they cannot act profitably, but this just isn’t true. As you’ll find on Ethisphere (https://ethisphere.com/) some of the world’s most profitable and respected companies are also some of the world’s most ethical companies too.
Ethics have to matter more than money in a well-functioning world. But don’t think that you have to forgo the first to have the second. That’s one of the best lessons that Aunt Shirley ever taught me.
"Things seen differently."