Author name: Mitch Solomon

Companies exist to solve customer problems. That’s the foundation of every successful business, and it should be the foundation of every sales conversation. Yet most sales training in industrial and operational technology companies starts in the same place: the product. That’s exactly where deals go to die. When salespeople lead with features, specs, and configurations, they reduce a strategic business decision to a technical comparison, and the consequences extend far beyond a single conversation. Deals stall as customers struggle to build internal consensus around value they were never shown. Opportunities shrink, margins erode into price comparisons, and salespeople lose access to the senior stakeholders who actually own the outcomes. Many deals don’t fail because the product was wrong. They fail because the customer was never given a reason to act.

Recurring revenue is among the most trusted signals of business health, and among the most easily misread. A book of contracted ARR with a high renewal rate is often seen as a powerful indicator of a business that is performing well. Yet these metrics often reveal something quite different, namely the durability of decisions made years ago under conditions that may no longer exist.

When people talk about why language models worked, they usually point to scale: bigger models, more compute, better optimizers. The truth is that the internet did most of the work in creating that scale. By the time GPT-3 arrived, decades of human writing had already been digitized, indexed, and made cheaply available. The training data wasn’t free, but it existed.

Industrial and operational technology companies face a uniquely difficult challenge in managing many aspects of their business, and culture is no exception. On the one hand, they must satisfy the rigorous expectations…

While the AI most people know runs in distant data centers, answering questions and generating images from the cloud, Physical AI operates where the action actually happens: on the device, at the edge, in the moment. It drives cars, pilots drones, stocks warehouses, performs surgery, and walks on two legs through factories and homes. Instead of mastering only language and pixels, it has to reckon with the stubborn unpredictability of the real world, often with milliseconds to decide and no round trip to the cloud to spare. Running intelligence on the edge, close to sensors and actuators, is what lets machines perceive and respond at the speed of reality. In short, Physical AI is the moment intelligence stops being a conversation in the cloud and starts being a presence in the room. Machines that don’t just think, but sense, decide, and do.

Companies that build and sell industrial and operational technology face a distinctive risk environment. Their products are often deeply embedded in customers’ critical workflows such as facility operations, fleet management, industrial process control, or transaction processing, which means that both the cost of failure and the value of trust are unusually high. They operate across long sales cycles, complex channel relationships, and demanding validation requirements, often selling to buyers who are slow to adopt and slow to switch. Many are navigating a fundamental business model transition, from hardware and perpetual licenses toward software, subscriptions, and outcome-based offerings, that requires them to take deliberate strategic risks at the same time as they manage the day-to-day risks of running an engineering-intensive business in a world being transformed by AI. That’s why at VDC Strategy we think carefully about how industrial and operational technology vendors should approach strategic risk, and why we think a more rigorous risk assessment framework is critical.