Innovation at the Edge: Lessons from Generali

with Danilo Raponi from Generali

Danilo Raponi, Group Head of Innovation at Generali, joins James Benham live at InsurTech Insights London 2026 to break down what insurance companies keep getting wrong about innovation — and what the rare successes actually do differently. From a blockchain art consortium that collapsed due to partner misalignment, to a three-question discovery framework for finding real pain points, to his prediction that humanoid robots will reshape insurance within five to ten years, Danilo brings first-hand lessons from building innovation capability inside one of the world’s largest insurance groups.

All Episodes Available On

The Failure Pattern Nobody Talks About

Insurance innovation conferences have a tell. There’s the big announcement. The press release. The consortium forming with three or four impressive partners. And then — nothing. Six months later, the project has gone quiet. A year later, it’s never mentioned again.

Danilo Raponi calls it the Keyser Söze pattern: the project goes up in smoke and everyone pretends it never happened. “Most people at events like this will only talk about their successes,” James noted during the recording. “Most humans just stop talking about things that don’t work out.”

Raponi disagrees with the impulse. His view is that the failures are more instructive than the wins — and that most of them share a common cause.

When he joined Generali, he inherited an active innovation project: a blockchain-based platform for art insurance. The consortium was impressive — a major insurer, a large bank, a stock exchange, and an auction house. The technology premise made sense. Blockchain could theoretically enable encrypted, immutable tracking of art transactions worldwide — transactions that are notoriously opaque. Each partner could layer services on top: insurance, lending, funding, authentication.

“People got excited about the technology,” Raponi explained. “With blockchain, we can monitor all the transactions happening around the world. But then, first problem: is this really what clients are looking for? Who are the clients?”

The second problem was worse. “The biggest issue that made everything blow up: misalignment between the partners. The insurance company was just looking to sell more art insurance. The bank was looking to do some lending. The auction house was looking to do more volume. Everybody had completely different objectives — which weren’t discussed and agreed before.”

The first round of funding went through because all the partners wanted to see the thing get built. But when it came to scaling, each partner suddenly had a different vision of what the company should become. They fought. It collapsed.

“Start with the problem, not the solution. That is the fundamental rule.”
— Danilo Raponi, Group Head of Innovation at Generali

Insurance Companies Are Not Tech Companies

One of Raponi’s sharpest arguments in this episode is one that sounds obvious until you start counting how many carriers have ignored it.

“Sometimes I think what insurance companies tend to do too much of themselves is to become almost tech companies, although they’re never going to make it,” he said. “They build, build, build inside, whereas there are people out there doing great things and you can just partner with them.”

His argument is structural, not cynical. If you are a carrier, your core competency is pricing risk, off-laying risk, and investing the float. If you are a TPA, your core competency is claims and policy administration. Both are already difficult, complex jobs. Adding a software development function to that creates overhead without competitive advantage.

“If you try to build it all yourself, you’re going to have a hard time getting everything done — because it’s not the business model,” he said. “As I always say to startups: the most important thing is focus. Do one thing, do it properly, do it well. Same still applies to large corporates.”

He applies the same discipline to Generali’s innovation lab: instead of running 100 experiments, pick two. Fund them properly. Put real people behind them. See how it goes.

The Florida property market, he said, is a cautionary case study for what happens when insurance companies move too fast without staying anchored in their core competency. “A lot of the markets in Florida improperly priced their property risk. They didn’t account for two things — hurricanes and lawyers. The latter being the one that really blew the market up.”

Carriers that expanded into Florida without deeply modeling the litigation environment found themselves paying out claims at a loss they hadn’t priced for. “It’s been really interesting to be around long enough to watch insurance companies go out of business. It actually happens.”

The Magic Wand Question

Much of the conversation returns to a single failure mode: building solutions before confirming anyone has the problem you’re solving. Raponi has watched insurtechs demonstrate impressive technology to heads of innovation at conferences like this one, receive a polite “yes, that looks interesting,” and then come back six months later with a six-figure price tag to find there is no buyer.

“They’re not even asking questions,” he said. “They’re just asking if you like what they built.”

The fix, he argues, is learning to ask questions — and knowing which questions to ask. He references a framework he calls the Magic Wand Question, which must always be the third question you ask, not the first.

The sequence:

What do you do?
What’s the most challenging part of your day?
If you could wave a magic wand and change one thing, what would you change?

“The third question is the most important question after you warm them up,” he explained. “That’s when you get a good insight into what the person is actually struggling with. If you ask the magic wand question cold, before the warmup, you’ll get a polished answer. After the first two questions, people have stopped being polite and started being honest. That’s when you’ll actually get the truth.”

“If you’d wave a magic wand and change one thing, what would you change? That’s when you’ll actually get truth.”
— Danilo Raponi

Generali’s innovation function applies the same discipline internally. Raponi’s team of 15 — plus ~30 innovation colleagues embedded in business units — maintains a portfolio of problems sourced from colleagues across the group. France has different pain points than the UK corporate and specialty team. The goal is to understand the problem deeply first, match it against the company’s strategic priorities, and only then go looking for solutions — or decide whether to build one in-house.

Robotics Is a Five-Year Problem

When James asked what Raponi is most excited about looking forward — beyond large language models, which he described as “now, not the future” — the answer was immediate.

“I think robotics. Robotics is going to be massive already in five years. People think about robotics as in 20, 30 years — but I think in five to ten years, we’re going to have metallic friends working around and helping us do stuff we don’t want to do.”

He’s not just watching from the sidelines. “Full disclosure — I’m an early-stage investor in a pretty serious robotics company,” he said. “I can tell you there are amazing things happening in the next two years. We are underestimating their impact.”

His warning to the insurance industry: “We are really not doing enough to understand what’s going to be the impacts — not only in the liability, which is the most obvious thing of these robots running around and doing things, but in just how are they going to change everything.”

The liability angle is intuitive: a humanoid robot injures someone, who is responsible? But Raponi’s point is that the harder questions are the systemic ones. How does widespread robot deployment change underwriting for workers’ comp? How does it affect property insurance when robots become part of the built environment? How does it change life and disability coverage assumptions built on human workforce models?

China’s manufacturing investment and Tesla’s Optimus program are the two evidence points he keeps returning to. The timeline most insurance leaders are using for scenario planning is off by twenty years.

Key Takeaways

Start with the problem, not the solution. The #1 failure mode in insurance innovation is technology-first thinking. Falling in love with blockchain, AI, or any shiny tool before confirming the problem is real leads to projects that disappear without a trace.
Focus is the superpower. Insurance companies trying to become tech companies dilute their core competency without gaining real advantage. The principle applies to seed-stage startups and 150,000-person insurers equally.
Partner alignment before money. Consortia collapse not from technical failure but from unaligned objectives that nobody discussed up front. Don’t skip the alignment conversation.
Robotics is a five-year problem. The industry’s planning assumptions about humanoid robot timelines are off by two decades. The liability angle is obvious — the systemic impact on every line of business is the bigger, less-asked question.