Nine Months After the SaaSpocalypse: What Michael Nadel Actually Learned

In late January, Anthropic released a new model. Then OpenAI. Then Meta. Then Google. Every two weeks, a new frontier model, a new headline about which industry it was about to take down.

By February, the market had made its verdict. Roughly $300 billion was wiped from software valuations in about two days, part of a broader $2 trillion reckoning that came to be called the SaaSpocalypse. The thesis was simple and terrifying if you ran a SaaS company: LLMs could vibe-code enterprise software, so traditional per-seat software was about to become redundant.

Nine months later, Michael Nadel, Global Head of Insurance at Simon-Kucher, sleeps a lot easier than he did in January. Not because the disruption didn’t happen. Because it happened differently than the panic predicted, and the difference matters enormously for how InsurTech companies should be thinking about pricing right now.

In Episode 171 of InsurTechTalk, Michael walked me through what actually separated the companies that got hurt from the ones that didn’t, why InsurTech vendors have flipped from offensive to defensive monetization strategies in the span of a year, and why the single most important pricing idea his firm has used for three decades hasn’t changed at all.

About Michael Nadel

Michael Nadel is a Partner and the Global Head of Insurance at Simon-Kucher, based in Chicago. He spent his early career at Accenture across banking, wealth and asset management, and capital markets, before moving to CNA Insurance, where he was part of the strategy and innovation team focused on modernizing legacy technology and partnering with InsurTech companies to solve carrier-specific problems. He joined Simon-Kucher in 2022, initially leading the North American insurance practice before being named Global Insurance Practice Lead. He holds an MBA from Northwestern’s Kellogg School of Management and a Master’s in Information Systems and Bachelor’s in Finance from Indiana University’s Kelley School of Business.

What Actually Happened in the SaaSpocalypse

The companies that got hit hardest, in Michael’s assessment, weren’t randomly selected. They shared a specific characteristic: they provided standardized, replicable workflow processes that relied less on genuine subject matter expertise. Project management tools. Point-to-point solutions that moved something from A to B without requiring deep domain knowledge in between.

The market’s underlying assumption was that internal tech teams could now reproduce that kind of software more simply. Nine months in, Michael thinks that assumption was incomplete in an important way.

Software Is Only a Small Part of What Enterprise Buyers Are Actually Buying

When a company buys enterprise software, they aren’t just buying the code. They’re buying security, uptime, stability, and a long list of operational guarantees that a vibe-coded internal solution doesn’t inherently provide. And even where AI has made teams incrementally better at building software, that doesn’t automatically make them better at maintaining it, hosting it, and supporting it at scale. If you weren’t a build shop before, becoming one overnight is harder than the panic assumed.

Benioff vs. McDermott: Two Statements That Aren’t Actually Contradicting

The public framing battle between Salesforce’s Marc Benioff (bullish, AI creates new revenue opportunity) and ServiceNow’s Bill McDermott (more cautious, “intelligence is being commoditized”) got treated in the press as opposing predictions. Michael’s read is more nuanced: both statements can be true simultaneously, and which one ends up mattering more depends entirely on how a given company positions itself in response.

Why Scale and Incumbency Matter More Than the Headlines Suggest

Large, entrenched platforms like Salesforce and ServiceNow benefit from something early-stage competitors don’t have: a large installed base deeply embedded in core business processes. That embeddedness is sticky. A CFO might want to reduce spend on a tool like that, but unwinding deep integration is genuinely difficult. For incumbents at that scale, AI capability becomes primarily an upsell and expansion opportunity within an existing account, not an existential threat.

The Evolution From Point Solutions to Platforms

A year ago at ITC, most InsurTech pitches Michael heard were narrow point solutions: an underwriting workbench, a single bolt-on to a policy administration system, solving one specific link in the insurance value chain. What he’s seeing now is a shift toward broader platform positioning: full-service intake through claims handling, agentic solutions spanning multiple workflow stages. Carriers are actively looking for IT landscape simplification and end-to-end outcomes, not another point solution to integrate and maintain.

From Offensive to Defensive Monetization in One Year

This was the sharpest and most immediately useful insight in the conversation.

A year ago, the dominant conversation among InsurTech software vendors was offensive: “I’ve built an AI feature into my product, and I want to monetize it as an upsell on top of my core software.” Now, Michael is hearing the opposite from those same kinds of companies: fear, and a defensive scramble to change monetization models before it’s too late.

The Underwriter Problem

The specific fear Michael described is worth sitting with directly: a company built its pricing on a per-user, per-month SaaS basis. The value delivered flows through a human underwriter using the software. But if that underwriter’s role is reduced or eliminated by AI, and the carrier still gets the same value from the underlying capability, the vendor’s revenue disappears even though the value they deliver hasn’t. The fix isn’t a better feature. It’s shifting the entire pricing metric away from counting human seats toward something aligned with value or outcomes, so the vendor gets paid regardless of who, or what, is actually using the product on the other side of the screen.

Why This Is Fundamentally a Moat Question

Michael connected this directly to competitive defensibility. Being deeply embedded in a workflow, a policy admin system, a claims system, a billing system, is one kind of moat. But if your pricing metric doesn’t align with the value being created or protect you when the underlying workflow changes, you don’t actually have that moat, no matter how embedded you think you are. Many companies are actively restructuring their pricing right now specifically to build that protection before it’s needed.

The AI Wrapper Problem

Michael was direct about a second, related risk: generic ChatGPT or Claude wrapper products offering no unique value are already getting washed out of the market. Beyond the competitive risk, there’s a margin risk few companies have fully modeled yet: if you haven’t priced in your actual AI compute costs correctly, you may be increasing output and increasing your own costs simultaneously, without capturing any of that value in your price. Michael expects a lot of carriers, and a lot of InsurTech vendors, to be genuinely shocked when they see their year-end AI infrastructure bills.

Price Is an Exchange of Value: The Idea That Hasn’t Changed in Three Decades

This is Simon-Kucher’s foundational pricing principle, and Michael’s articulation of it is worth remembering precisely: price is not a number, it’s a measure, the same way a foot is a measure of distance. What you’re actually doing when you price something is aligning that measure to the value a customer perceives.

The Thumb-in-the-Air Problem

Michael described a recurring pattern with fast-growing Series C companies: years of R&D, a fully built sales team, genuine product investment, and then, when asked how they arrived at their price, essentially a guess. Working backward from an already-built product and an already-set price to figure out what’s wrong is far harder than starting with customer research before you build anything. Price functions as the tip of the spear of your entire strategy. It signals whether you’re a premium product, how you should be interpreted relative to competitors, and what kind of customer you’re actually built for.

The “Five Customers, Not a Thousand” Problem

One of Michael’s clearest diagnostic questions for early-stage companies: how many of your thousand customers are actually paying? When the answer comes back as five, the real customer count is five, not a thousand. Free users solve neither the chicken-and-egg problem of needing traction to raise capital nor the deeper problem of not knowing what people will actually pay for.

Where CFOs, CMOs, and Sales Teams Actually Collide on Pricing

Pricing projects, in Michael’s experience, tend to be unifying rather than divisive internally, mostly because they usually surface real margin leakage that everyone in the room wants fixed. But alignment isn’t automatic. Simon-Kucher often runs what Michael calls a “goal-strat exercise” at the start of an engagement specifically because executive teams frequently discover they don’t actually agree on the objective: are you maximizing revenue, profit dollars, or profit margin? Those are different goals that can require different pricing strategies, and misalignment on that question is often more revealing than any customer research finding.

The sales organization carries a particular kind of risk in any pricing change: they know exactly who they sell to and how they pitch, and a shift in pricing metric can upend their entire target account list and pitch strategy, sometimes with serious consequences for a technology vendor selling into the insurance industry specifically.

Intelligence Is Commoditized. Judgment Is Not.

Michael offered a genuinely memorable, concrete example of where the line between AI capability and human judgment still sits.

Drafting his fantasy football team under time pressure, he asked an AI tool for a pick recommendation. It suggested Josh Jacobs, a star player, without accounting for the fact that Jacobs had been suspended from the league that same day. With five seconds left on the clock, the tool caught its own error and reversed course. Michael’s point: AI can search and synthesize information faster than any human ever could, but it still requires significant human oversight, instruction, and judgment to catch exactly this kind of failure, the kind of failure that, in an insurance context, could mean wrongly denying or approving a claim based on incomplete context.

The honest caveat: this is anecdotal, not statistically rigorous. But the pattern Michael described, needing additional verification agents and human oversight layered on top of AI output, reflects something genuinely unresolved about where AI reliability actually stands today.

ITC Vegas 2026 Masterclass Preview

This is Simon-Kucher’s third year running a version of this masterclass, and its second as an officially embedded part of ITC. Attendance has grown from roughly 40 people in a booked conference room to nearly 200 last year, with high expectations for continued growth this year.

This Year’s Structure

• Section one: AI’s impact on monetization for both InsurTech software companies and services businesses, presented alongside a Simon-Kucher colleague who focuses on software broadly, not insurance specifically, to bring cross-industry perspective
• Section two: A panel of carrier and vendor executives discussing how buying behavior has actually changed, what carriers are evaluating when purchasing software today, and how they think about paying for AI use cases specifically
• Section three: The impact of changing monetization strategies on sales organizations, and how to keep them effective through that transition

Who Should Attend

Michael’s honest answer: primarily anyone selling technology or services to a carrier or broker, at any stage from early idea to Fortune 500 public company, because the specific way AI impacts pricing strategy differs meaningfully by company stage. An early-stage company spending a much higher percentage of ARR on marketing than an established public company faces a fundamentally different pricing calculus.

The Blockbuster Warning

Michael’s closing framing for the masterclass: many InsurTech companies today have a genuinely good product that could work well in the future, but if they don’t adapt their monetization model to the AI transition correctly, they risk becoming Blockbuster, a company with a fine product that simply didn’t see the shift in how customers were about to procure value coming around the corner.

The Closing Answer: The Industry Needs to Advocate for Itself

Asked what the industry should be talking about more, Michael’s answer moved away from AI and pricing entirely. His genuine frustration: insurance does a poor job advocating for itself publicly. The industry absorbs significant public criticism, visible on social media, about not acting in customers’ best interests, while very few carriers are willing to publicly defend individual cases or make the broader case for the value the industry provides. Michael’s honest belief: the world genuinely could not function without insurance. It enables people and businesses to take risks that would otherwise be impossible. He would like to see the industry get meaningfully better at telling that story to the outside world.

Key Takeaways

• The companies hurt hardest in the SaaSpocalypse were standardized, replicable point solutions requiring minimal subject matter expertise, not enterprise software broadly
• Enterprise buyers pay for security, uptime, and stability alongside the software itself, capabilities a vibe-coded internal solution doesn’t automatically replicate
• InsurTech vendors have shifted in one year from offensive AI monetization (charging more for an AI feature) to defensive monetization (protecting pricing from a shrinking human seat count)
• Aligning your pricing metric to actual value delivered, not to a human seat that AI might reduce or eliminate, is becoming the real competitive moat
• Price is fundamentally an exchange of value, not a number; working backward from an already-built product to figure out pricing is far harder than starting with customer research first
• AI has commoditized intelligence and information retrieval speed, but not judgment, the human oversight layer that catches context AI misses remains essential
• The insurance industry does a poor job publicly advocating for its own value, and would benefit from telling that story more assertively