Dirty data will determine the new digital divide.
A side note in the October 2019 issue of Inbound Logistics highlights that – according to a Telematics Benchmark Report – most (fleet) organizations lack a data driven approach to running their business. While 86% of the respondent companies use telematics, only 23% make use of big data analytics for strategic decisions, and 36% are still reliant upon manual processes for hiring forecast requirements.
In other words, these companies may know where their trucks are in real time, but are struggling to otherwise make critical business decisions using reliable data.
The consequences of bad data can be summed up in a sidebar article titled “Dumb data has a cost” in the August 2019 issue of Logistics Management magazine. According to Dun and Bradstreet’s report “The Past, Present, and Future of Data”, approximately 20% of businesses have lost a customer due to bad data. And 22% of businesses reported their financial forecasts were inaccurate due to bad data.
According to Monica Richter, chief data officer at Dun and Bradstreet: “Whether leaders are exploring AI or predictive analytics in their supply chain, clean, defined data is key to the success of any program and essential for mitigating risk and grown business.”
Artificial Intelligence (AI), blockchain, Internet of Things (IOT) … these are the buzzword technologies that are swirling around right now. A (long?) while ago it was XML (which never killed EDI) and radio frequency identification (RFID, which never displaced the ubiquitous barcode). The facts are that the new technology investment costs could not be justified, especially with the realization as to the massive scale of the changes that would have to be incurred. Even with Walmart pushing RFID, the initiative stalled.
It is not that any of these are bad technologies. Nor should they be necessarily classified as solutions looking for problems. After return on investment, the biggest blockade to implementation is dirty data, not just the enterprise seeking to implement the technology, but across its supply chain partners. Master data management is both an internal and an external function.
What I predict is – and will continue – to happen is that poor data governance practices will doom enterprises and create a new type of digital divide that defines classes of businesses: those able to embrace new technologies versus those unable to embrace new technologies. As new technologies become mainstream necessity technical tools, enterprises unable to embed these technologies within their business operations and software platform portfolios will be unable to competitively meet market challenges, and will ultimately lose out to their more able and stable competitors and newcomer startups.
Not all new technologies will be necessary for an enterprise to implement. But certainly enterprises that continue to think that they can roll the dice with dirty data are fooling themselves into complacency.
Companies need to remain vigilant to insider threats.
As reported in the August 2019 issue of DC Velocity magazine, a 2018 study produced by Through Transport Mutual Insurance (TT Club), a British transport and logistics industry insurance provider, highlights that cargo theft is primarily an inside job.
Criminal organizations recruit carrier company insiders to provide information on cargo routes, destinations, cargo loads, and access to information technology systems.
Favored commodities are food and beverage (19%), followed by alcohol and tobacco (15%), consumer products (15%), electronics (7%), and then apparel and footwear (5%).
The Association of Certified Fraud Examiners (www.acfe.org) recognizes the following means of reducing fraud:
· The more that there is a perception of detection, the less there are incidences of fraudulent behavior due to the fear of getting caught.
· The separation of responsibilities decreases the incidences of fraud.
· Job swapping, whereby employees change roles – and therefore responsibilities – on occasion. (This would necessitate updates to software application user rights and network security profiles.)
Where there are multiple steps required to perform a task, checks-and-balances (including data governance) must be implemented where it makes sense to catch illicit activity but being mindful of not inhibiting throughput.
Restricting the ability to extract data from software applications is something else to consider. We tend to exist in an open environment where almost any data that a user can access can be easily saved in spreadsheet format and therefore downloaded or emailed. While this can be helpful in promoting more effective office communications and data analysis, it can also be very useful in perpetrating illicit behavior. Software system (e.g. Enterprise Resource Planning, ERP) security should be more restrictive in what it allows individual or classes of users to do with data from a reporting and download standpoint.
Organizations these days tend to focus on the external threat: viruses, phishing, hacking. They are right to do so, and need to remain vigilant of threats that seek to worm their way onto their networks and penetrate their software systems and sensitive data files. However, organizations cannot forget that they are also threatened by the disgruntled employee who is swooned or swayed by the outside entity who entices the insider to illicit behavior.
Donald Cressey, the criminologist who coined the term “while collar crime”, developed the fraud triangle which states that there are three components to fraud: opportunity, rational, and pressure. When a person is under enough pressure and has the opportunity and can rationalize the behavior, fraud will occur. Companies must reduce the pressure, remove the opportunities, and discourage the thought of the commission of the illicit behavior. This takes leadership and a cultural commitment at all levels of the organization.
The median value of cargo theft was $19,000 in Asia, $60,000 in Europe and North America, and $77,000 in South America. The preferred method of cargo theft was road transport (versus other methods I assume such as air, rail, or water).
40-year-old technologies are the backbone of supply chain success.
Perhaps shockingly, three 40-year-old technologies are the backbone of supply chain success. If your company does not have a successfully implemented Enterprise Resource Planning (ERP) system with integrated Electronic Data Interchange (EDI) and cannot successfully produce item and carton barcodes, you are pretty much out of luck.
Granted, the science behind these technologies has advanced over the decades: laser versus charged-coupled device scanners give quicker barcode reads; the Internet is faster than dial‑up modems; ERP systems have more functionality and interoperability than before, plus there are now different hosting options. But the basics still hold true: these three technologies are at the foundation and the core of supply chain success.
XML – with its dozens of variations – never killed EDI. RFID (radio-frequency identification) didn’t put the ubiquitous barcode out of business. And inasmuch as blockchain seems to be having some isolated project success, I am not holding my breath for it to kick EDI to the curb anytime soon.
I fully understand that any new technology requires hype to generate excitement and create mass adoption which in turn lowers prices and allows standards to shake out.
But reality soon sets in when practicality, cost, and return-on-investment are considered, as well as priorities such as just trying to get it right with what you’ve got.
Companies that I have been engaged with struggle to just get 40-year-old technologies up‑and-running well internally and externally with some of their supply chain partners such as contract manufacturers and third party logistics providers, let alone the thought of introducing newer technologies that would serve absolutely no use or benefit. These companies lack the skills, training, and leadership to just get these decades-old foundational technologies on board and integrated with good business operations practices.
The result is that their supply chain performance suffers … often times terribly. And while the company’s technology leaders look towards untested or buzzword-bound products and services to solve their woes, they fail to understand that the foundational technologies remain ERP, EDI, and barcoding. There is no need to recreate the wheel: this stuff already exists, and it actually works quite well.
Retail, pharmaceutical, medical products, publishing, electronics, marine, government, apparel, footwear …. I have helped companies in a variety of industries, and the supply chain requirements remain served by the same three foundational technologies: ERP, EDI, and barcoding.
These three same technologies play big roles in both of my books. These supply chain systems – and the transactions they output – are used to analyze supply chain performance of a company’s vendors (and can be used to analyze its own internal supply chain performance too), and can be used to analyze for possible illicit behavior (a.k.a. fraud).
If your organization is struggling with the selection and/or implementation of any or all of these three supply chain technologies, especially as it impacts business execution, please contact me for help.
Industry 4.0 requires success with Industry 3.0 first.
An interesting article in the April 2019 issue of Inbound Logistics magazine identifies 5 insights to Industry 4.0.
According to a September 2, 2018 Forbes.com article by Bernard Marr, Industry 4.0 can be defined as “computers are connected and communicate with one another to ultimately make decisions without human involvement.”
Industry 3.0 was “the adoption of computers and automation” according to Mr. Marr, with humans still implied to be involved. Industry 4.0 apparently nixes the need for people to be part of the equation.
But as the Inbound Logistics article noted: “the biggest barrier to implementation is poor communication.” This statement should not be relegated to just Industry 4.0, but Industry 3.0, Industry 2.0 (“mass production and assembly lines using electricity” per Mr. Marr), and Industry 1.0 (“mechanization through water and steam power” per Mr. Marr).
In fact, this statement applies to any operations improvement or software implementation project. And sure enough, in the modern industrial era, the “poor communications” identified in the Inbound Logistics article are those between operations and information technology, two traditionally siloed areas that seem to have difficulty coming together towards a common goal.
(The article failed to include the accounting and finance area(s) as a typically siloed group whose perspectives are not always in alignment with those of the operations folks and technology area. And let’s not forget about the sales and marketing teams who tend to have unique priorities of their own.)
Industry 4.0 cannot be successful if it is not built upon a solid foundation of a fully functional Industry 3.0, both internally and externally to the organization. You may desire to outsource a task – whether to a vendor or to a machine – but you cannot outsource that responsibility. Before you can even think about automating business processes and the backend decisions that cause them to come alive, you had better darn well ensure that there was – and is – effective communications that got you to where you want to be and monitor the situation to ensure you stay within the acceptable parameters of what you consider to be normal.
Just because there are data communications based on data-driven automation does not alleviate the fact that there may still be too much manual intervention to be effective. Companies today struggle to implement Enterprise Resource Planning (ERP) systems and seamlessly interconnect with customer and vendor supply chain partners via Electronic Data Interchange (EDI). Data quality is poorer than it should be. Data governance gaps – operational and technical – allow the infiltration of damaging data to infect the organization. How can the automation of Industry 4.0 be expected to function upon a foundation of less‑than‑perfect data?
As the old saying goes: To err is human, but to really mess things up takes a computer. And an automated system left to its own devices, unchecked or monitored after-the-fact, relying on bad data, will only cause destruction.
Bridging the gaps between functional areas and supply chain partners takes enterprise-level understanding, the ability to fit data into software and understand how it will support the business, the foresight to visualize how business processes can be achieved with software tools, and the skills to be a relationship champion both internally and externally to the organization. Ensure your leap from one industrial version to the other doesn’t become an industrial disaster.
Are ERP project successes exaggerated?
People overstate and exaggerate all of the time. It is done innocuously without ill intent. I admit I have difficulty with quantitative measurements such as length, weight, temperature, and distance so asking me my opinions on those matters should be taken with a grain (okay, I know how much that measurement is) of salt. When asked how far or how fast, I am likely to unknowingly over-state or under-estimate.
(And yet as a fencer, I have very little problem with effectively using distance to my opponent offensively and defensively. Just don’t ask me the length of that distance!)
Exaggerations – when done innocently – are, I think, just a natural part of the human condition. Exaggerations can also be done depending upon the basis for comparison. For example, based on research by the University of Pennsylvania School of Nursing, senior citizens (e.g. 75 years and older) with chronic illnesses are generally rating their health as excellent or very good. The study researchers were curious about this discovery. The respondents informed that as long as they were able to do what they liked and enjoyed, even with their illnesses, they considered their health to be “good” or better. Within the context of their age and their ability, the study participant’s responses were not quite exaggerations but need to be framed within their overall state of being.
However, when they are done with a directed and nefarious purpose, exaggerations are really more appropriately branded as intentful misdirections or, more directly, lies.
There was an interesting sidebar article in the March 2019 issue of Inbound Logistics magazine that caught my attention. Titled “ERP Drives ROI”, the very short piece presented statistics on the success rate of ERP implementations. (Given that one of my core specialty areas is Enterprise Resource Planning – ERP – systems, I was particularly drawn to this information.)
The article noted that 67% of responding manufacturers and distributors rated their ERP implementations as successful or very successful over other answers such as partially successful, not very successful, and failed.
The article stated what should be obvious reasons for success: people, processes (which I take to mean having defined processes), and having top management support.
Likewise, the article stated what should be obvious reasons for failures: inadequate business processes and lack of project planning.
But the article also noted that inasmuch as many ERP projects meet their budget and timeframe goals, success expectations may be exaggerated. At the end of the project, when the budget has been expired and the timeframe has been reached, top management may exaggerate ERP success by reframing the initial expectation, e.g. leaving out functionality that was supposed to be initially implemented as part of the original go-live plan. So, can the ERP project really be considered a success if less was implemented for the same budgeted amount and the same original timeframe?
Why would executives, and even middle management, exaggerate in this manner? Two answers come to my mind, given that I have witnessed this first-hand on numerous occasions: pride and politics.
In the twenty ERP and financial systems projects I have been involved in, I have witnessed numerous project failures (such as timeframe overruns, go-live failures, and budget excesses) due to the reasons stated in the article, but also due to pride and politics of the actors involved. Large-scale projects are no time for protecting head count, creating siloed departments, favoring an unqualified vendor, or resisting change.
Certainly, it is tough on the ego to admit defeat, but truth is better than lies. When exaggerations become egregious to the point where they are no longer innocent extremes, you have compromised your morals, values, principles, and ethics.
I know that executive management has to sometimes act as the cheerleader to rally the troops. But staff are smart, and they will quickly begin to know when exaggerations have turned to excrement. This affects company morale and senior management credibility. My message to senior management is to prepare for project success from the beginning.
Instead of risking the success of an ERP project from the outset, ensure that you take all of the necessary precautions to be successful from the start. Don’t exaggerate about the project team you think you have; be honest about the help you need at the get-go so you’ll have a better chance of fulfilling your promises at the end.
We need more echoes in the canyon.
I recently saw the music documentary film “Echo In The Canyon” (https://www.echointhecanyon.com/). If you are a fan of ‘60s music, this film is a must-see. Hosted by Jakob Dylan – who is the splitting image and replicates the sound of his famous father, Bob – the movie is about the music scene in the Laurel Canyon, California area between 1965 and 1967. Profiling musicians from and bands including The Byrds, The Beatles, The Beach Boys, Buffalo Springfield, The Monkees, and The Mamas and the Papas, the film is a tribute to some of the most iconic music that has outlived the generation from which it was born.
But beyond a celebration of music and the composing musicians, the film is an inspirational lesson or two about so much more.
Listening to the memorable songs sung by current-day performers and hearing the stories by iconic musicians should be enough for any music lover or aficionado. The tales of parties between famous band members and how musicians would freely share musical riffs that they created but didn’t want with other musicians who took them and developed them into now-famous songs to create music history is nothing short of shear entertainment.
It was the openness to free-form collaboration without fear that someone else – a colleague or competitor band – would benefit more that enabled each musician to gravitate to the ideas that truly worked best for them and turn those strums and plinks into hits. It was about the ability to work with – not against – another musician regardless of whether they were in the same band as you were or not. It was this non-musical harmony that resonated throughout the movie and throughout each musician’s interview. Any competition, if it could even be called that, was really more inspiration. There was no real one-upmanship. There was cross pollination of ideas that were tailored to individual talents.
As I have written in VAR Insights magazine (https://www.varinsights.com/search?keyword=%22norman+katz%22) and as I repeatedly promote in my supply chain vendor compliance consulting (see the June 18, 2019 Forbes.com contribution by Kate Vitasek in which I am profiled:
https://www.forbes.com/sites/katevitasek/2019/06/18/abusive-compliance-practices-hurt-retailers-and-their-vendors/) there are ample opportunities for collaboration but, I think, rare instances of people and organizations actually doing so.
Software and hardware resellers should be partnering with independent consultants who bring depth, experience, and executive interaction to the table. Independent consultants know how to sell the value proposition rather than selling the technology. Article after article I read implores the reseller away from the commodity sale and towards the value proposition, even recommending direct-hiring of people with the independent consulting experience. But fear of splitting or sharing the revenue and the misguided belief that independent consultants are competitors to resellers prevent this collaboration from regularly occurring.
In my 25 years of supply chain vendor compliance, I have only witnessed the widening gap between the customer/buyer entities (e.g. retailers) and their supplier/vendor sellers. The Internet has leveled the playing field to a great extent, shocking retailers who have finally come to terms with the fact that they have disenfranchised their vendors and cannot fully private label their way out of the disloyalty hole they have dug for themselves. Brick-and-mortar’s failure to fully grasp and merge the omnichannel business model has only widened the chasm and created more supply chain chaos within their own companies, let alone the spillover to their vendor communities. With numerous retailers completely out of business resulting thousands upon thousands of stores closed and jobs lost, I am still waiting for anyone to fully address the fact that the lack of true supply chain collaboration is a significant cause of retail distress and failure.
What those hippie-dippy musicians were actually teaching us, decades later, is that success for one equates to success for all, even if it comes at different times and at different levels. The lesson is that collaboration is not a weakness but a core strength. Your bandmates – your teammates – are your driving force, even if sometimes they drive you a little crazy in the process. And if someone else can benefit from something you created but can’t use, just let them have it without strings attached.
Even through the quarrels and the fights, the bandmates and musicians never really compromised on their collaborative efforts when it came to the music.
From software projects to operational improvements to supply chain collaborations, these efforts are best solved by creating echoes in the canyon. When it comes to better business execution, what we could use are more collaborative-minded people to join together in the band. The results would be business hits.
Retailer bags return-fraud thief red-handed.
Only in superficial South Florida I suppose.
39-year old socialite Meghana Rajadhyaksha was caught by retailer T.J. Maxx for return fraud when it was discovered that she was purchasing expensive designer handbags that retailed for between $1000 and $2000 from the retailer’s online site but then returned a counterfeit copy in their place.
T.J. Maxx caught on to this scheme and then informed the U.S. Secret Service because part of the U.S. Secret Service’s mission is counterfeit investigations.
The socialite from South Florida could certainly afford to purchase – and keep – the expensive handbags she was buying. A transplant from Detroit where she owned a hotel, her husband is an orthopedic surgeon. Caught up in Miami’s social scene, our fraudster enjoyed hobnobbing with Miami’s social elite. This likely explains why she needed expensive handbags and couldn’t be caught dead using the same one more than a few times.
According to the Miami Herald, the Miami society maven’s scheme involved 69 handbags from such famous brands as Fendi, Dolce and Gabbana, and Gucci that were worth a total of approximately $135,000. The fraud spanned a 17-month timeframe whereby the goods were shipped to her at her $1.7 million dollar home.
The scheme unraveled when the retailer noted the fraudster’s buy-return transaction pattern. Simple enough. Examining the returned merchandise, which the fraudster even tagged with the retailer’s own and original merchandise tags from the real items, the retailer realized that the returned goods were fakes. (No information on where the clever copycat bags were sourced from.) The retailer then set up the fraudster by intercepting the goods she was purchasing and marking them with invisible dye so that if they were returned, as the pattern suggested, the expectation was that the original – and now marked – goods would be returned. However, as anticipated, unmarked goods were returned instead of the real items as the retailer suspected they would be.
Not only was T.J. Maxx defrauded, but the fraudster also sold some of the real handbags that should have been returned to T.J. Maxx through consignment seller RealReal, realizing a net profit of $11,000.
The socialite made no attempt to conceal her love of high-priced handbags, as she was photographed with them at the various social gatherings she attended.
The fraudster’s mistake was certainly to commit the crime in the first place.
However, our criminal was realistically identified with good data analytics. The high rate of orders and returns over time would be a metric that any online or brick-and-mortar retailer should be looking for. This fraudster should have spread her fraud out among several or many retailers over time which would have made the fraud less likely to be noticed because the number of orders and returns over a time period would have been reduced for each retailer. Even more so, she should have kept a handbag or two to reduce the chances of being caught. But our fraudster got greedy, and she certainly got caught.
According to her defense attorney, the socialite schemer has accepted full responsibility, is attempting to make whole the retailers for their losses, and is in professional therapy voluntarily seeking help. She still faces possible prison time.
Even amidst the blizzard of transactions that occur in retail, or in any business, it is still possible to use same transactions that operate a business to detect and reduce supply chain fraud. You don’t need to be a big retailer or a large corporation to engage your company’s software systems and the transactions that they produce to harness their fraud-fighting power.
If you want to learn more about the potential of your company’s software systems and the transactions they are producing to detect and reduce supply chain frauds, or to solve other competitive challenges, please contact me.
Retail store closings continue into 2019.
The March 2019 issue of Inbound Logistics magazine had two sidebar articles about the struggles of brick-and-mortar retail stores.
The first is titled “48 Hours, 465 Store Closures” that encapsulated the store closings of JCPenney, Gap, Victoria’s Secret and Footlocker combined all within a 2-day period.
The second is titled “No Light At The End Of The Retail Tunnel” which reports that store closings may not be slowing down into 2019 given that 2,187 stores have closed so far this year between the shutterings of Gymboree, Shopko, and Charlotte Russe. Keep an eye on a second possible bankruptcy for Payless ShoeSource this year.
These closings come on the heels of 5,524 store closings in 2018, as the second article reports. What is to blame, the article states, is online retail growth, flat or declining store sales, and rising interest rates.
Yet experts continue to agree that consumers would rather touch-and-feel before they buy. It would be interesting to know the rate of returns at the online retailers consumers are supposedly turning to as they shun physical stores. Likewise, how are these return policies, which probably favor consumers, eating into profitability? Returns costs not only requires postage but also inspection and putaway – translation: people – not an inexpensive addition to operational overhead.
So, one has to wonder why consumers are no longer interested in going into those stores. Staid, tired old merchandise? Unfriendly help? Lack of new and interesting products? Humdrum experience?
For decades, retailers have been building walls between themselves and their vendors with draconian, ill-conceived, poorly communicated vendor compliance programs. I contend that what retailers have created has now come back and exacted revenge upon its creator. Retail vendors lack the loyalty to care who they sell to, and are branching out to sell direct to consumers in some cases. Where retailers complain that they lack supply chain visibility, it is because vendors lack the loyalty to care to fully commit to these abusive, and in all likelihood temporary, relationships.
Retailers continue to invest in the latest cutting-edge technology for their customers, but lack the commitment to the supply chain vendors to accept their responsibility and take the leadership to correct the problem. Retailers have failed to figure out that omnichannel – which is all about shopping and not about shipping – was their own business model chaos that they forced upon their vendors who, for decades, have been successfully shipping to stores, distribution centers, and drop-ship direct to customers long before the term “omnichannel” came in to being.
If retail executives are going to get serious about saving their stores, it is going to start with an introspective look at their supply chain vendor compliance programs that define their vendor relationships, old and new. It is going to require a hard look at how they treat their vendors and whether these relationship definitions and parameters live up to the lofty words and ideals of their corporate values and mission statements … because I’ll tell you that they realistically don’t. It is going to take real leadership to navigate the necessary change to make amends for the sins of the past and present and chart a better course in creating the real collaboration that should be defining the essence of what supply chain is all about. Or we can all just sit back and watch as more stores shutter and more companies close.
Abusive vendor compliance programs are affecting retail store success – here’s how.
Between my books, speaking engagements, and writing assignments, my international presence just keeps growing.
Just last month in March, my 6,700 word, peer-reviewed, statistics-support article “How abusive vendor compliance programs are affecting retail store success” was published in Volume 1 Number 3 of the Journal of Supply Chain Management, Logistics, and Procurement by Henry Stewart Publications.
I have written about how vendor compliance is an overlooked opportunity by retail executives to save their stores from being shuttered. I outlined and then provided detail in my book as to how poorly implemented vendor compliance programs cause disruption, inhibit innovative products from reaching consumers, turn competitors into commoditized businesses, increase costs, and alienate supply chain vendor partners.
And now supported industry statistics from leading supply chain magazines, research organizations, and university studies, I examine how the retail industry is in a state of paralysis when it comes to its own supply chain execution. Despite its own self-serving supply chain accolades, the retail industry’s relationship with its vendors it pitifully poor and is a root cause problem for its woes.
For an industry that seeks the most advanced artificial intelligence and data analytics to use to engage its customers, it is still relying upon 40-year-old technology for its vendor engagement supply chain in barcodes and Electronic Data Interchange (EDI). Draconian vendor compliance programs are so muddled, muddied, and in some cases questionably legal that it is hardly any wonder why vendors have no retailer loyalty and are engaging the Internet to go direct to consumer, bypassing retailers all together.
From removing the reasons to enter a store to failing to understand that omnichannel is about shopping, not shipping, so don’t blame your vendors for your own flawed siloed business models, retail executives continue to point the finger of blame in the wrong direction for supply chain disruption, and now it is snapping back hard.
My message to retailers is: You can and need to do better, now.
I have been deeply engaged in the technical and operational aspects of supply chain vendor compliance since 1993. I authored the first and exclusive book on the topic – “Successful Supply Chain Vendor Compliance” – in December 2015. And I am now honored to be recognized by my peers in being published in such a prestigious journal with an editorial board comprised of professionals from recognized global corporations.
For those of you who wish to subscribe and receive a copy, please go to:
I have a few article reprints that I would be happy to mail out within the United States. (If outside the US, I can send via UPS or FedEx if you provide me your account number.) If you are interested, please contact me with your mailing information.
Artificial Intelligence is subjective, not objective, and this will cause some concerns.
I attended a technology panel discussion in December 2018 and one of the trend topics discussed was artificial intelligence (AI). Naturally, as this is one of the hot technology trends seeing greater application use. We see AI being pushed into our consciousness via IBM’s commercials for its Watson AI product.
One of the panelists discussed the ethical concerns about using AI to assist physicians in making medical decisions. An AI tool can be used to help doctors quickly cull through the growing encyclopedic collection of reference materials related to a pharmaceutical drug or ailment to assist a patient with a path to recovery. However, this is where there is pushback because people are concerned that the AI tool is replacing physician’s insights and oversight in the decision-making process.
And this got me thinking.
What if one company that has an AI tool to help physicians in such a way uses the AI from IBM, and another company that has an AI tool to help physicians in such a way uses the AI from another company, e.g. Microsoft. Even if both AI tools are feed the exact same data, because they are AI, my reasoning is that they could – and naturally would – end up with different results sometimes, just like people do. As such, the different AI tools, depending on their creators, would at some point, eventually, advise the doctors differently when asked the exact same question using the exact same data.
You can test my theory today if you would like using two popular Internet search engines: Google and Bing.
As someone who often does a lot of research, I have found that these two search engines will return different results even when I search on the exact same word or term. As a researcher, I appreciate the value of using both search engines in having the different perspectives and retrieving the different results, as it benefits my research greatly.
Try it for yourself. Open up Google in one browser tab and Bing in the other, and select a search word or term and check the results. (The browser you use will not matter.) Certainly, these search engines are pretty close to being like AI tools, so I think that this is a fair test and comparison.
As AI becomes used to help with data analysis and embedded into more and more complex devices – think autonomous vehicles – the selection of whose AI may become exceedingly important as to whether AI is right for the use, and how much AI influence there should be in the decision-making process.
There might come a time where multiple AIs from different companies are embedded in a machine or device and – if they do not agree – the human is the final decision maker. This creates a type of consensus where people still have ultimate control and no single company’s AI is the preferred product. Inter-connection might be forced like it was back when ODBC (Open Data Base Connectivity) was created, essentially when the software community had to allow their competing systems to speak to each other because of the needs of the business community. We might see AI develop the same way.
When people speak of AI, they speak of it as it is a single thing or entity, and it is not. AI is a product, just like any other piece of software, that is being developed by different – and competing – companies. Whose product wins out may not be a factor of whose product is better: it may simply be a situation of whose product has got the better-funded marketing engine behind it. So, an end-product with embedded AI may not be using the most advanced AI tool in the marketplace, just the AI tool whose company had the better marketing or was able to ink the better partnership agreement.
Given that AI is bound to provide subjective, and not objective, results, what concerns does this cause for the end user who will be reliant upon the results of the AI’s decision which, ultimately, could have had as its foundation, a preferred business deal?
The legal, moral, and ethical considerations make my head tingle. And that is my real intelligence at work.
Don’t let video kill the rock-star recruit.
For those of you not engaged by music trivia and who cannot relate to the catchiness of the title of this month’s newsletter, it is a reference to the song title of the first video played by MTV when it made its on-air debut on August 1, 1981. “Video Killed The Radio Star” by The Buggles was the first video played by MTV.
In December 2018 I had the opportunity to interview for an interim project manager position at a public university. A grant-funded role, I did not know how long it would last, but I never know how long any of my assignments will be. The university used a video interview service where the candidate would (supposedly) be presented with a question, have time to think of an answer, and then record the answer. A computer and web camera were the tools needed. Okay, this is a little different but I could see the practicality of it. The execution, however, was a disaster and failure from a project management standpoint.
The video interview third-party software provider had technical glitches which they acknowledged, forcing my first and second interview attempts to fail. The company only provides telephone support from 9:00 AM to 5:00 PM Central time Mondays through Fridays, (I am located in the Eastern time zone, as is the educational institution), even though interview candidates were informed they could take the interview at any time day or night. The video interview had a five-day deadline that ended on a Sunday at midnight from the date that I received the interview request. Otherwise, if help was needed outside of support hours and days, only a help desk ticket could be submitted.
The technical glitches eventually seemed localized to my choice of web browser, as my third attempt using Internet Explorer rather than Chrome alleviated the problems. However, no prior instructions as to which preferred browser to use were provided. In fact, the educational institution’s best guidance was for me to go and use someone else’s computer or perhaps go to a library and use a computer there. Really? For a video interview of undetermined time and length where I had to ensure a quiet place? Are you kidding me?
As a candidate, I was providing screen snapshots of the error messages I was receiving to the educational institution. I was acting as technical support and project manager. Why didn’t the educational institution establish functioning computers at their location and have all candidates come in to record their interviews instead of burdening candidates to be debuggers and run around finding alternate computers to use?
Once I was able to finally access the interview, I found the setup to be contrary to qualifying job candidates, at least for a project management role. Was this a Hollywood screen test or an interview for a project manager role? There were supposedly only seven questions to answer, with one minute to consider each question before recording one’s answer. However, the questions were multiple parts, with one question having four parts to it. The number of takes option was disabled, therefore the candidate could not re-record a bad video take upon stumbling during the 3-minute answer timeframe. The video interview should have been limited to one, short, concise question per take, not one multi-part question per take. And the candidates should have been allowed to record up to two takes per question to ensure satisfaction with the video-recorded answer.
Not even the structure and length of the interview was known up front. Information about the interview format and length would have been appropriate to provide to the candidate prior to the interview itself, such as in an introduction email.
And the premise was obviously false: a multi-part question is not the same as a single-concept question, especially when the different parts of a multi-part question are unrelated from each other.
I am not opposed to the use of technology, nor am I against new methods of doing things. What I absolutely do not agree with is when poor project management and a lack of foresight and planning create a disadvantage to achieving success and what should be the stated project goals and objectives. I also spent four hours of my time debugging the first two interview attempts just to try and accomplish the third.
As a friend of mine who is a life-long human resources executive pointed out, the problem with a video interview is that it can be labeled as discriminatory because the candidate’s image is visible. At one time it was thought that a candidate should attach a photograph to their resume, with that practice quickly scuttled for this reason. An automated interactive telephone interview to initially screen applicants, where candidates receive a call into a service, are asked a recorded question, and then record a response – something I have also recently experienced – is a much better option because the company has the resume and only a recorded voice, no candidate image attached to any candidate profile. And this was done right: one brief question at a time, and it felt like a natural conversation to me on the telephone.
Organizations are having endless conversations about wanting the best and brightest “rock‑star” candidates, but processes like the video interview that I described above will only alienate and discourage quality candidates from bothering to consider companies that cannot get their hiring processes organized. More egregious are the ridiculous psychological online games that companies attach to hiring processes in a belief that they are predictors of a person’s personality profile. Having sampled some of these myself, I don’t need to be an expert to tell you that relying on these is about as scientific as online dating. If this is how your company is qualifying candidates, you’ll basically get what you deserve.
For companies who are disgruntled about the people they are hiring, resentful with new hires who fail to show up, angry with new employees who leave their jobs after being on board only a short while, dissatisfied with personnel who don’t live up to expectations, maybe placing the blame on the person you hired is short-sighted: perhaps it was the poor process that produced the exact results that you received that needs an examination.
Time to set a new corporate course for the New Year.
On November 1, 2018, 94,000 Google employees worldwide walked off away from their jobs and protested against the company rather loudly. They were mad – not just upset – at their company’s poor handling of sexual abuse allegations against executives.
Google’s original tag line that expressed its company’s ideal was “Don’t be evil”. Employees apparently did not believe the company was exactly living up to that vision as it paid out excessive amounts to executives charged with questionable behavior.
Google – now part of the parent company Alphabet – follows an updated mantra: “Do the right thing”.
Dictionary.com defines “ethics” to be “a system of moral principles”. Great. Going further, Dictionary.com defines “moral” as “of, relating to, or concerned with the principles or rules of right conduct or the distinction between right and wrong”.
Ah, easy enough. Ethics is really just doing what is right and not doing what is wrong.
But companies don’t really do this.
From customer service that is really risk management to deceptive pricing entangled in bundled products or services that don’t really represent what the consumer wants to pushing products or services that the customer doesn’t actually need to gleaning information for “customer service” (a.k.a. marketing) purposes, companies engage in unethical behavior every second of every minute of every hour of every day with little concern for their conduct.
Led by executives who are actually people who make the decisions to engage in such wrong practices, these leaders would be shocked to discover that this unethical behavior is actually costlier to their companies than doing what is right. Case in point is that is costs five to ten times more to acquire a new customer than to retain an existing customer. So, when customers lack loyalty and continually switch brands or expand their horizons and try new competitors, forcing companies to increase their marketing because data analytics tells them to, they are only causing their own chaos. When it takes seven customer service agents between the call centers and the “Office of the President” special staffs to handle upset customers, these unnecessary operating costs are because the company at its core is not doing what is right.
The Google employees were protesting more than just how their employer was handling mischievous executives; they were also fed up with certain forced arbitration clauses in their employment contracts among other issues. The were demanding that Google act ethically across a number of employee-impact areas.
Companies have stakeholders in their customers, suppliers/vendors, employees, and investors. Ethical behavior needs to be at the forefront of all stakeholder considerations. Inasmuch as profits are a primary driver, they are not ever, and no longer, the only important consideration and should never eclipse ethical behavior. In fact, the more ethical a company behaves, the cleaner their operations will actually be, and the more loyal their stakeholders will be too.
Shockingly, the people in charge forget that people are impacted, e.g. their own friends, family, and neighbors. Not even inside connections can always help to untangle the labyrinth-like web that is weaved by companies who have grown not to be too big to fail, but rather, are actually too big to cohesively function.
No company is perfect. But ethical companies have found that they can win without compromise.
For information on some of the most ethical companies in the world and how they out-perform the rest, go to: https://www.worldsmostethicalcompanies.com/