(Image credit: IIR/Gemini.)
Insurance carriers are facing a familiar problem with increasingly significant implications: technology decisions made to solve one problem often create others. Cloud adoption offers a good example. Many insurers moved workloads to cloud providers because the cloud was perceived to be cheaper. In many cases, it wasn’t.
That does not mean the move was wrong. Cloud environments can deliver important advantages in disaster recovery, business continuity, flexibility and scalability. But those benefits were not always the original goal. When the expected result is cost reduction and the actual result is a different operating model—with new dependencies, new governance requirements and new cost dynamics—the decision needs to be understood in a different light.
Every decision has consequences, which is clearly the point. A problem or opportunity is recognized, a plan is developed, resources are deployed and action is taken. In the postmortem phase, actual results can be compared with expected results, and a new set of decisions can be framed.
Simple enough in theory. In practice, however, the unexpected consequences can be highly consequential. In a worst-case scenario, the cure can be worse than the disease. A mentor once shared with me that surprise outcomes are generally bad because they reflect a failure to plan. Clearly there are surprise endings that turn out well—Teflon and Sticky Notes come to mind—but only after accepting that the original goal of those science projects had failed.
The Cost of Lost Optionality
Insurance technology leaders must think this way because the stakes are high. Hardware and software providers have sharply increased prices in recent years, often recognizing how easily vendor lock-in can be monetized. This kind of “greenmail” can be avoided when optionality is built into architectural plans, but that often runs counter to near-term cost reduction efforts. In other words, Mister Goodwrench may have been prophetic with the “pay me now or pay me later” mantra.
A smaller example recently reminded me of the same pattern. In an effort to get ahead of the end of support for Windows 10, I retired an old computer in favor of a new one. The new machine was faster in most ways, but it tripped over Wi-Fi compatibility issues with a pandemic-era router. A new router led to a mesh network upgrade. Eventually the problem was solved, but in an expensive and time-consuming way. The punchline was that Microsoft then extended the life of my Windows 10 license. I was undoubtedly better off for the long term, but the path there was far more complicated than expected.
AI and the Pressure Downstream
The explosion of AI-related activity creates a more important version of the same issue. Armed with a cadre of new tools, organizations can now generate code at extraordinary speed. That can be powerful, but it also creates new pressure downstream. Code still needs to be tested, integrated, secured, managed and maintained. Unless the funnel for those downstream functions expands to match what is coming in at the top, a quick productivity gain can become a source of corporate swirl and paralysis.
The analogy is nautical. Dramatically increasing the number of people on one side of a carefully balanced boat does not make it more efficient. It can swamp it. In the worst case, it sinks.
None of this is an argument for becoming a Luddite. Technology advances and operational change can produce profoundly good outcomes. For insurers, cloud, AI and modernization remain essential to competitiveness. But planning never goes out of style, and managing accelerated exposure to unintended outcomes is part of the price of admission as the industry races toward 2030.
That means insurers need more than enthusiasm for new capabilities. They need architectural optionality, realistic cost models, governance capacity and a disciplined view of operating risk. They also need something that has become harder to find: time for executives to think critically about major decisions before momentum makes them difficult to reverse.
That time has sharply declined over the past two decades, perhaps as an unintended consequence of an always-on, always-connected world. For effective executives, however, this is a solvable problem. The first step is accepting that there is a problem—and that we know what it is.