The Baruch Group

Call Directions Contact the organization

The Baruch Group Disaster Preparedness . . . Business Continuity . . . Enterprise Risk Management Offers -
Consulting services for businesses and government agencies.

Tenacious and conscientious protection of employees, assets, and the reputation of the enterprise
Establishes productive collaborations among the numerous stakeholders
Designs cost-effective response and recovery protection from natural and person-caused disasters

10/23/2026

Managers Need to Get Better at Killing Digital Products
by Evan Shellshear
https://buff.ly/PTo6xjx
October 14, 2026
Connect Images/Getty Images
Summary. Companies often hesitate to sunset digital products because they fear customer backlash. But a study of roughly 360 product and service shutdowns over nearly two decades suggests that fear is exaggerated: About two in five generated no measurable public reaction, and only about a third of those that were noticed provoked a negative response. The research reveals that companies should separate two questions: Will anyone notice the shutdown, and, if they do, will they be angry? Factors companies can control—including communications—primarily affect visibility, not customer sentiment. Backlash depends more on whom the shutdown leaves behind: Declining products may retain a small but intensely loyal user base particularly prone to protest. Companies should therefore calibrate their shutdown efforts to likely visibility, focus resources on helping highly attached customers transition, and be candid about why a product is ending. Done thoughtfully, sunsetting can free resources for more promising products without creating the reputational damage companies fear.close
In a 2019 global study of more than 1,600 product managers, product end-of-life was the lowest-rated of 15 core product-management skills. The ability to develop an effective end-of-life plan was the second-weakest of 74 individual capabilities measured.

Why is ending a product so difficult? Ask executives why they hesitate to sunset a digital product and you’ll hear some version of the same worry: backlash from customers. Everyone working in the digital world can name a discontinuation that triggered petitions, angry press, and customer defections. That memory does real damage. It encourages companies to keep zombie products alive long past their usefulness—or wrap every shutdown in elaborate damage-control machinery on the assumption that catastrophe is the default outcome.

But is that fear justified?

To find out what actually happens when companies end digital products, I studied roughly 360 documented shutdowns of digital products and services spanning nearly two decades, coding each one for how the public reacted. As a part of the data collection, I researched why the product or service was shut down, what happened to the staff, what type of notice was given and much more. The data drew from public digital records—news outlets, social media, blog posts and more—and covered information along over a dozen dimensions on each shutdown. The goal was to shed valuable insights on what factors are most important to focus on when winding down a product or service.

The findings overturn most of the conventional wisdom about how to end things. They also happen to be reassuring.

Finding #1: The fear is greatly exaggerated.
Two data points force us to rethink the whole problem. First, roughly two in five of the shutdowns I examined produced no measurable public reaction at all: no coverage, no organized complaint, nothing a determined search of the public record could later find. Second, even among the endings that did register publicly, only about a third turned negative.

Put those together and the feared brand-damaging backlash is not the rule but the exception. Most endings, even of real products with real users, simply pass.

That single fact should change how much money and anxiety companies pour into the average shutdown. But the more useful discovery is why some endings blow up? The reason is not what companies think.

The mistake almost everyone makes is treating a shutdown as one event with one outcome: Did it go well or badly? It’s actually two questions in sequence, and they behave completely differently:

The first is: Did the ending register with the public at all?
The second, which only arises if the answer to the first is yes, is: Given that people noticed, did they turn on you?
Conventional shutdown advice—i.e. how to phrase the announcement, how much notice to give, and which channels to use—blurs these together, as if the same moves that get an ending seen also determine whether it’s forgiven.

They don’t.

Separating the two questions is the most valuable thing a company can do before planning a shutdown, because the levers you control act overwhelmingly on the first question and barely touch the second, which brings me to the next finding.

10/22/2026

McKinsey Technology Trends Outlook 2026
September 15, 2026
Report, Sixth edition
The 14 tech trends
By Michael Chui, Roger Roberts, and Tanguy Catlin
https://buff.ly/ZP28aQ8
Which frontier technologies matter most for companies in 2026? Our annual report highlights the latest technology innovations, developments, and talent trends and their potential impact on business and society.
The technology story of 2026 has moved off the screen and into the physical world. Innovation is accelerating in the power grids and chips that underpin the data center boom; in the intelligent robots that embody AI; in the agentic systems discovering new chemical compounds; and in the launch pads sending thousands of satellites into orbit.

AI needs energy to scale. That’s one reason energy technologies alone drew nearly $200 billion in investment in 2025, among the highest capital influx in any technology domain. And spending on AI infrastructure doubled in a single year. These developments show that the defining questions today are not only about what technology can do. They are also about who can build the hardware and assemble the skilled workforce to deploy AI in the real world. At the same time, huge leaps were made in cybersecurity and software development—illustrating that AI is accelerating the digital frontier, too.

McKinsey’s Technology Trends Outlook 2026 examines 14 technology trends that define 2026, expanding our coverage from last year to include two new fast-emerging domains: agentic software development and AI for scientific discovery and engineering. For easier navigation, we group the trends into three broader categories: AI revolution, compute and connectivity frontiers, and cutting-edge engineering. The lines between these domains are blurring, and much of the innovation is happening in the gaps.

The report is based on extensive quantitative analysis. We measured metrics on innovation, investment, interest, talent demand, and organizational adoption (see sidebar “Research methodology”). It also draws on qualitative interviews with business leaders, policymakers, investors, and experts across sectors and geographies. The technologies included in this report were selected based on their rate of innovation, as well as their growing applicability and potential to influence how we live and work (see sidebar “How we chose the trends”).

Four of this year’s trends are AI specific, but AI also underpins and amplifies the other ten. AI has become an accelerant in robotics, creating a new class of autonomous bots capable of interacting with the world around them. AI is reshaping software development, making coders many magnitudes more productive. AI is advancing scientific discovery, helping life sciences companies discover new drugs many times faster than before. And AI is optimizing the foundations of technology infrastructure itself, changing cybersecurity operations, and catalyzing new innovations in semiconductors. These are just a few of the ways AI is shifting the sands beneath our feet.

How we chose the trends
Beyond the AI ecosystem, many of this year’s trends are being shaped by large-scale economic, geopolitical, and workforce transitions. Scientists and engineers are discovering new possibilities in therapeutics, diagnostics, genomics, and brain science. Robotics, mobility, and immersive-reality technologies are converging to reshape how humans interact with digital and physical systems. Sustainability technologies are transforming how energy is produced, managed, and used—providing potential solutions to meet surging power demand from data centers. And quantum computing continues to advance, giving a glimpse into how it could soon solve problems that would otherwise be infeasible.

This breakneck pace of change comes with challenges. Organizations are racing to deploy AI at scale without any proven road maps. They have workforces that need upskilling, legacy systems that need updating, and networks exposed to ever-evolving security risks. They face shortages in energy, talent, and capital. Overcoming these roadblocks requires far more than just deploying technology. It requires rewiring operational models from the inside out.

10/22/2026

Three reasons why AI is not conscious and perhaps never will be
Large language models can mimic thought and feeling with uncanny skill. But evidence from neuroscience suggests that intelligence alone is not enough for subjective experience.
By Anil K. Seth, Ned Block & Giulio Tononi
Artificial-intelligence systems such as large language models (LLMs) can exhibit behaviour that seems strikingly intelligent. This has led many people, including some experts, to wonder whether these systems could have some form of consciousness — the capacity for subjective experience. However, we think that existing AI systems are not conscious, and that, along current trajectories of development, they are not likely to become conscious. Greater clarity about whether AI could become consciousness will be crucial as society grapples with the challenges posed by this increasingly powerful technology.

Can AI ever be conscious? The question stems from a misconception

Several prominent scientists, including biologist Richard Dawkins and AI pioneer Geoffrey Hinton, have argued that AI systems might already be conscious. And earlier this year, AI company Anthropic in San Francisco, California, published research claiming to have identified computational signatures of a form of consciousness in its Claude model1.

Whether AI models can ever be conscious is not merely a philosophical question. It has practical implications. For instance, if AI systems were conscious, they might experience distress and pain, similar to how people do. Then should countless new conscious entities that might experience suffering be allowed to proliferate? Perhaps not surprisingly, some people are already calling for ‘AI welfare’ and ‘AI rights’ to be taken seriously2.

Conversely, if AI systems only seem to be conscious and lack actual inner experience, then granting them rights or protections could hinder efforts to govern and control them at precisely the time when oversight is urgently needed. Following this path would draw attention and resources away from beings whose consciousness is not in any doubt.

Looking beyond humans and our closest relatives in the animal kingdom, it is still hard to be certain about what does, and does not, have the capacity for consciousness — be it insects, AI or the creations of synthetic biology, such as laboratory-grown brain organoids3,4.

Why we need mandatory safeguards for emotionally responsive AI

This uncertainty has prompted some researchers to advocate a precautionary approach: if there is a reasonable chance that AI systems could be conscious, then society must take anticipatory action and give those systems some form of protection against potential suffering. This, the argument goes, would avoid reprising moral errors of the past — for example, the mistreatment of non-human animals, or surgery on preterm human babies without anaesthesia. An extreme version of this idea is that society should reallocate resources from addressing the welfare concerns of present-day humans to lessen the potential suffering of trillions of future AI descendants.

In our view, uncertainty is not the same as knowing nothing, and it does not mean that all possibilities are equally likely to occur. Decisions about what is moral must weigh both the consequences of false negatives — failing to recognize consciousness where it exists — and the costs of false positives — attributing consciousness where none exists. In the case of AI, we think that false positives are much more likely to occur than false negatives are. If so, a precautionary approach would be misguided.

10/21/2026

Do You Know What Your Customers Value?
https://buff.ly/EmNf3qI
Not according to the data. Here’s how to fix that. by Vikas Mittal, Alessandro Piazza, Sonam Singh and Ashwin Malshe

From the Magazine (November–December 2026)
Gabriele Galimberti
Summary. Many executives mistakenly believe that because they’ve spent heavily on gathering customer feedback, monitoring competitors, and analyzing market trends, they understand what customers want. But the authors’ seven-year study of 72 companies and additional research involving thousands of customers and firms reveal that companies routinely misidentify the factors that drive customer value. That happens because leaders base their customer strategies on intuition, recent events, or competitors’ moves rather than scientific evidence. Not only that, but they tend to change their strategic priorities frequently, which leads to organizational clutter, confused customers, and wasted resources. The solution is to use data and statistical models to identify which customer value drivers have the greatest impact on sales and profitability, align employees around a small number of priorities, and eliminate low-value activities. Companies that adopt this disciplined approach can simplify strategy, strengthen customer loyalty, improve productivity, and generate superior financial returns.

Most companies work hard to get close to customers. Managers also regularly compare how their product offerings stack up against the competition, so it’s not surprising that executives are confident about their knowledge of customer needs and whether they’re satisfying those needs better than their rivals do.

What we saw was that executives did indeed spend a great deal of time listening to customers, monitoring competitors, and looking to spot industry trends. But we found that rather than using a scientific approach to interpret the information they gathered, they routinely took cognitive shortcuts. As new factors surfaced and different executives focused on different ones, companies’ strategies for winning customers shifted from year to year, confusing customers about the underlying value propositions the firms were offering. As customers switched to competitors, sales dwindled. Meanwhile, the costs of pivoting caused margins to shrink, customers didn’t get what they wanted, employees were burdened by the need to work on multiple initiatives that added little value, and shareholders earned average or below-average returns.

In this article we present the findings from our research and offer an explanation rooted in behavioral science for the phenomenon we describe. We’ll then outline a five-step framework that will help executives sharpen their focus on what customers really value, using examples from companies we’ve studied or worked with to demonstrate how it works.

What Executives Get Wrong
If companies don’t know what their customers really want, it’s not because they lack data. At 93% of the companies we studied, executives spent four to 12 hours a week, on average, gathering direct customer feedback, monitoring competitors, and analyzing trends. As we’ve noted, nearly all of them felt confident that these activities ensured that they were focused on customer needs.

Unfortunately, it turned out that most of them also relied on intuition and personal experience to interpret that data. Only 2% of executives and employees in our surveys said their companies had formally identified the features in their value proposition that customers most appreciated (such as product quality, purchasing experience, and after-sales service) and had assessed their relative importance to customers. None had quantified how much each of these “value drivers” actually affected revenue or margins.

Even when companies did identify and compare value drivers, there was little internal agreement about which were the most important. When we asked executives in each company to rank its drivers of customer value, fewer than 10% of their responses agreed. When the employees of each company were asked what customers valued most about their firm’s offerings, virtually all listed the same things: quality (100%), service (100%), sustainability (100%), competitive price (98%), AI leadership (98%), customer experience (96%), communication (91%), innovation (91%), safety (88%), and project ex*****on (87%). (Note that many of the firms we studied were B2B businesses, which explains the high rankings of AI leadership and project ex*****on.) When asked which was their company’s priority, 96% of the employees responded “all of the above.” Only 4% provided a rank ordering.

10/21/2026

Your Company Needs a Plan for the First 72 Hours After a Cyberattack
by Harpreet Sidhu, Jason Lewkowicz, Rouzbeh Hashemi and Yusof Seedat
https://buff.ly/jJVYTdb
October 12, 2026
Audshule/Stocksy
Cyberattacks now move too quickly for companies to assume they can simply restore every system and resume normal operations. The organizations that recover fastest prepare to run the business while parts of their technology environment remain unavailable or untrusted. They define a “minimum viable enterprise”: the essential outcomes, decisions, applications, and manual fallbacks needed to protect customers, employees, compliance, and revenue during the first hours and days of disruption. Leaders should identify that core before a crisis, assign decision rights, map the dependencies that could halt it, maintain a clean recovery environment, and regularly test their plans under realistic conditions. Cyber resilience is less about restoring everything quickly than knowing what must keep running when everything else goes down.

Most companies operate with one of two assumptions about cyber disruption: that strong enough defenses will prevent it, or that a well-written business continuity plan will manage it when it happens. Our research, drawn from a global survey of 1,000 senior executives across 20 industries in 13 countries, suggests both assumptions are dangerously wrong.

AI-enabled adversaries have sharply accelerated the pace and scale of cyberattacks, with attacks increasing by 89% year-over-year in 2025, according to a recent CrowdStrike report. As autonomous, AI-driven techniques become standard practice, attackers can scan environments, identify vulnerabilities, and move laterally across systems in minutes. That speed has compressed breakout time—the interval it takes for an attack to spread beyond its initial entry point—from 62 minutes in 2023 to just 29, drastically reducing the window for response before disruption cascades across the enterprise. As AI models become more capable, the response window will keep narrowing, turning today’s difficult or expensive techniques into routine attacks and leaving organizations even less time to detect and contain a breach. At the same time, the cost of discovering exploitable vulnerabilities has already collapsed to almost nothing.

Yet according to our research, most (73%) of executives still believe predict-and-control security approaches are sufficient and persist with a strategy that experience disproves. While 81% expect that if a cyberattack struck today, their critical systems would be restored within 10 days our analysis of real-world cyber disruption cases places the actual average at three to six months. And while almost every company now has a business continuity plan, the evidence suggests many of these plans are not built for the realities of cyber disruption: business interruption remains the single largest driver of cyber losses, accounting for more than half of all cyber claims.

The gap between recovery expectations and reality exposes a structural misunderstanding of resilience. Traditional business continuity plans often assume that companies can restore operations quickly from fundamentally trustworthy systems.

Those that recover fastest start from a different premise: in a cyber crisis, the business may need to operate before its technology environment can be fully trusted. Instead of attempting to restore everything at once, they identify and preserve the minimum core operations needed to sustain trust, maintain continuity, and support decision-making, even when essential platforms like phone lines, Wi-Fi, internet, communication networks, and information flows remain unavailable or compromised for extended periods.

These companies pinpoint what we call their “minimum viable enterprise”: the smallest set of decisions, applications, and processes needed during extreme events. By preparing this essential core, they can recover much faster than companies that struggle to determine what to recover first, who has the authority to make decisions, or what manual fallbacks to activate when digital tools fail.

Protection and Resilience
After a global IT outage in 2024, for example, one U.S. airline we advised was largely operational by the next day because it had already defined the essential operations, decisions, and fallbacks needed to keep moving during disruption. AI can make this approach more dynamic and precise, helping organizations continuously map their environments and dependencies, identify the systems, data, and connections that support the essential operating core, and keep that understanding current as the business evolves. Combined with zero-trust controls that restrict lateral movement and limit the blast radius, this allows recovery to focus on what matters most rather than everything at once.

10/20/2026

When risk leaves the building: Are organisations really managing third-party risk? https://buff.ly/01JTRsx
As businesses become increasingly reliant on external technology providers, outsourced services and interconnected supply chains, managing risk can no longer stop at the organisation's own boundaries. By Siva Padaychee, Head of GRC and Business Assurance at e4
Issued by e4
Johannesburg, 18 Oct 2026

For many years, third-party risk was largely viewed as a procurement responsibility, focused on supplier selection, contractual agreements and pricing. Today, however, the conversation has shifted considerably. As businesses become more dependent on cloud platforms, software as a service (SaaS) offerings and outsourced services, third-party risk has become a fundamental business continuity concern.

The challenge is that organizations are increasingly reliant on providers that are deeply embedded in their operations. Moving away from an underperforming vendor is no longer as straightforward as it once was, particularly where proprietary technology, complex integrations and lengthy migration processes are involved. When a critical supplier experiences a disruption, the consequences can extend well beyond the immediate service failure, potentially resulting in security vulnerabilities, reputational damage and financial losses.

The risks we don't always see
One of the most overlooked aspects of third-party risk management is the exposure created by fourth parties, the suppliers and service providers that support our own vendors. An organization may have conducted extensive due diligence on its primary supplier, but what happens when that supplier experiences a service interruption because one of its own partners has failed?

This interconnectedness means that a business can be exposed to risks well beyond the relationships it directly manages. Other common blind spots include change risk, where a supplier makes technology updates without adequate communication, potentially disrupting client operations, and concentration risk, where multiple critical business functions depend on a single provider.

In the latter scenario, a disruption affecting one supplier could have a widespread impact across the organization . These dependencies are not always immediately apparent, particularly in businesses where individual departments manage their own vendor relationships without a consolidated view of the organization's overall exposure.

Due diligence cannot be a once-off exercise
A common misconception is that completing a thorough supplier assessment at onboarding is sufficient to manage third-party risk. In reality, no business relationship remains static indefinitely. Suppliers evolve, ownership structures change, services are modified and the broader risk environment shifts.

Prudent third-party risk management therefore requires ongoing oversight, with the frequency and depth of reviews informed by the criticality of the service being delivered. While cyber security credentials remain important, organisations should also consider a supplier's financial stability, business continuity and disaster recovery capabilities, regulatory compliance practices, geopolitical exposure, environmental, social and governance (ESG) policies, and governance structures.

The objective is to understand not only whether a supplier can deliver a service today, but also whether it can continue to do so under changing circumstances.

10/20/2026

The decisions that matter before, during, and after a cyber-attack
Building cyber resilience: introduction
The cost of disruption is rising, Resilience starts before an attack
Oct 16, 2026
https://buff.ly/erSLTt5
Authorities around the world are paying closer attention to how businesses, their boards, and management teams plan for cyber resilience. Drawing on our work across hundreds of cyber incidents globally, this series examines how multinational organizations prepare, respond, recover, and withstand scrutiny across an increasingly fragmented regulatory landscape.

In brief
The information needed in a serious incident is often spread across different teams and systems and has rarely had to come together at speed before. Where that work has been done in advance, leaders can make decisions much sooner.

Once an attack is under way, many of the most consequential recovery choices are about which parts of the business to bring back first. These are commercial judgments. Technology teams provide the picture and carry out the plan, but the call on priorities sits with business leadership.

Restoring operations rarely ends an incident. Regulators, investors, and counterparties then look back at how the organization prepared and how it handled the response, including what it said while the facts were still emerging.

Resilience starts before an attack
Even organizations with mature, sophisticated security controls can be successfully attacked. For that reason, resilience is less about avoiding every attack than about continuing to operate when safeguards are compromised. Much of that depends on decisions made long before an incident.

For multinationals, those decisions carry an added layer of complexity. Regulatory expectations around cyber resilience continue to expand and diverge between jurisdictions. The infrastructure on which a business runs may be global, but accountability to regulators remains local. Structures that work across the business also need to meet the expectations of authorities in different markets.

Rather than survey every regime, this series uses developments in key markets around the world to show where expectations are heading. The same shifts are visible in many other jurisdictions, and the patterns described are likely to apply well beyond the examples given.

The cost of disruption is rising
The financial consequences of a serious incident can be considerable. The average data breach now costs businesses almost USD5 million, and more than twice that (USD11.5m) in the United States. However, the most significant costs often come from interrupted operations, unfulfilled sales, customer churn, and reputational damage as a result of encryption events. When IBM asked organizations that had reported an attack in 2026 whether they had recovered completely, fewer than half (42%) said they had.

That exposure does not stop at an organization’s own systems. Big businesses depend on extensive networks of suppliers (including cloud providers, software vendors, and outsourcing partners) that can span dozens of jurisdictions. The World Economic Forum reports that 65% of large enterprises see third-party and supply chain vulnerabilities as their greatest challenge in building cyber resilience. A single compromised supplier can affect many of its customers at once.

Attacks are also getting faster. CrowdStrike reports an 89% increase in attacks by adversaries it classes as AI-enabled. Once inside, attackers take an average of 29 minutes to move beyond the first system they compromise. More often than not, the way in is through people and accounts rather than technical flaws.

Palo Alto Networks found that in 65% of its incident response cases, initial access came through identity-based techniques such as social engineering and the misuse of legitimate credentials. An attacker using a genuine account can look much like an ordinary user. The result is less warning and less time for executives to make their first decisions, which puts more weight on what is already in place before an attack begins.

10/19/2026

AI and the end
of loneliness

AI companions may soon cure loneliness, says psychologist Paul Bloom: a gift for the isolated and the elderly, but a danger for the young, who need to experience the hard work of real relationships. So what happens to a generation raised on chatbots that never need an apology or get bored of a story? Bloom pitches a stranger, more hopeful solution, one that lets us lean on AI without ever losing our humanity.

AI for loneliness is like a powerful painkiller: it can alleviate suffering, but can sometimes come at a terrible cost. “Think about the many millions of elderly people, many in institutions, who go for days and nights without talking to anybody, in perfect agony of isolation,” says Bloom. For these people with no one else to turn to, a properly attuned AI chatbot could be a godsend. For everyone else, sidestepping the sting of social failure means losing out on the chance to grow.

Loneliness pushes us to seek out social connection, but today’s AI doesn't encourage that behavior. “Imagine a teenage boy who never learned to cut short his monologues because his AI companion never got bored,” says Bloom. The solution isn’t as simple as turning off a chatbot’s sycophantic nature, because its next version — which performs pushback, but still lets you win — could be even more corrosive.

The best scenario would be to create imperfect companions. “Companions that are sufficiently imperfect that we ultimately abandon them to deal with people,” says Bloom. While there might not be money in building models people intentionally outgrow, he’s hopeful there are people “who care enough to do it anyway.” https://buff.ly/Fi6hKDu

Address

San Marcos, CA
92079

Telephone

+16502792556

Website

Alerts

Be the first to know and let us send you an email when The Baruch Group posts news and promotions. Your email address will not be used for any other purpose, and you can unsubscribe at any time.

Contact The Organization

Send a message to The Baruch Group:

Shortcuts

Share