The Quiet Shift to Outcomes as A Service

This article is by Sanjay Nursing, Insurance Specialist, SAS 

For years, insurers have tried to modernise by expanding the tech stack. New fraud systems, claims platforms, data tools, pricing engines, and increasingly, AI pilots. All of them have value, but when you talk to UK insurance leaders today, there’s a recurring theme: for all the investment in software, the impact on combined ratio and operational performance hasn’t always kept pace.

Too often the value of modernising the tech stack has landed in the software, not in the results. This isn’t because the technology is wrong, it’s because the way insurers buy and use technology needs to evolve (and it has been quietly doing so).

The industry has spent a decade paying for access – licences, logins, modules, but today’s pressures demand something more direct. Rising fraud, inflationary claims costs, climate-driven volatility and increasing regulatory scrutiny mean insurers are now far more focused on measurable improvements, not just additional tools.

This is where the emerging model of Outcomes as a Service (OaaS) comes in.

From paying for tools to paying for results

At its core, OaaS flips the typical SaaS model on its head. Instead of paying for software access, insurers directly pay for the outcomes the software delivers – whether that’s reduced fraud, faster claims processing, more accurate pricing or reduced leakage.

The idea has gained momentum because AI has reached a point where outcomes can be measured and monitored in real time, making it possible to align commercial models directly to performance.

Fraud is one of the clearest examples of this. Recent research shows fraudulent insurance claims cost insurers an average of £84,000 per case in the UK. It’s a stark reminder that traditional fraud systems were built before the scale, complexity or organised criminal patterns of today’s market.

When each incident carries that level of financial weight, it becomes far more compelling to work with partners whose fees are tied to avoided losses rather than user licences. AI already enables real-time anomaly spotting, so linking pricing to results feels like a natural evolution.

Claims are following a similar path too. Across the UK and Europe, insurers using AI-driven claims assessment and routing are reporting faster decisions and better customer experiences. Multiple studies indicate claims processing times can be reduced by around 30–50%. That’s a measurable shift. And measurable shifts lend themselves to outcome-based contracts around speed, accuracy and satisfaction, not just usage.

Pricing is the next logical area. Only about a quarter of P&C insurers say they have advanced modelling capabilities (just 27% in one global survey), so many therefore see AI-based, dynamic-pricing approaches as a key next step for staying competitive.

Major European insurers such as AXA are deploying or piloting more personalised motor insurance pricing models (for example usage-based tariffs and behaviour-driven rate-adjustments) rather than relying solely on traditional fixed annual premiums. In those contexts it becomes increasingly plausible that insurers will shift from licensing for pricing engines to paying for measurable improvements, such as lower loss ratios or reduced leakage.

Why AI makes OaaS possible now

Five years ago, this shift wouldn’t have been realistic. Today, it is.

This is largely because of the growth of agentic AI. These systems don’t just analyse data; they can take safe, governed actions such as routing cases, spotting fraud and assessing documentation. They handle repeatable, rules-based decisions at a scale and speed humans can’t.

And as soon as AI can consistently make those decisions, outcomes become trackable. As soon as outcomes are trackable, they can be contracted.

Importantly, none of this removes the need for human oversight or strong governance. In fact, it strengthens it. The Financial Conduct Authority (FCA), the Prudential Regulation Authority (PRA) and other regulators have made expectations around explainability, fairness and monitoring very clear.

This aligns closely with what we see in SAS’s own research. Our recent insurance study with Economist Impact found that leaders see the global protection gap, estimated at US$1.8 trillion (approx. £1.4 trillion), as both a challenge and an opportunity. But critically, they emphasised that modern data and AI governance is now the foundation for any high-impact transformation.

Put simply, OaaS is not a shortcut around governance, it depends on it.

Where OaaS is likely to scale next

While fraud, claims and pricing are the early movers, several other areas are ripe for outcome-based models.

Customer support is one of them. Virtual assistants already handle a significant proportion of insurance queries in some markets, and while UK customers still expect human escalation for complex issues, AI’s role in simpler interactions is growing fast. Metrics like “queries resolved” or “accurate routing” lend themselves naturally to OaaS-style contracts.

Underwriting is another. Many processes are still slowed by manual data collection and static workflows. As AI-driven risk assessment becomes more robust, insurers will be able to measure improvements in speed, hit rates and loss ratio – and align commercial terms accordingly.

Catastrophe response is also beginning to show early signals. With global insured CAT losses reaching US$137 billion in 2024, and the UK seeing more frequent flood-driven spikes, real-time surge modelling is becoming more valuable. Forecast accuracy and operational readiness are outcomes that can, in time, support performance-based contracts.

A quiet evolution, not a dramatic overhaul

OaaS isn’t going to replace SaaS overnight, and it shouldn’t. Insurers still need strong platforms, reliable analytics, high-quality data and tools they control. And in fact, OaaS depends on these foundations. You cannot buy outcomes unless your analytics and governance are already solid.

What OaaS does represent is a shift in emphasis. Rather than paying for systems that might deliver value, insurers can begin paying for the outcomes they actually achieve. In a market dealing with fraud spikes, inflation, climate volatility and regulatory pressure, that alignment is more than appealing.

And insurance isn’t alone in this shift

Other industries are already experimenting with versions of OaaS.

In cybersecurity, for instance, several vendors now offer limited breach-response warranties, where they provide financial support if an incident occurs – an early step toward outcome-aligned pricing. In logistics, route-optimisation platforms increasingly position their value around measurable fuel and efficiency savings, with some exploring pricing structures linked to those gains. And across cloud and managed services, providers are gradually shifting from simple licence or consumption fees toward more value-based models that reflect the efficiency or performance improvements they deliver.

These examples show that OaaS isn’t a leap into the unknown, it’s a continuation of a broader trend across technology-driven industries. Technology and the systems around them are and will always evolve.

This transition won’t be loud or sudden. As mentioned, it’s already happening quietly in specific use cases where the economics make sense. But over the next few years, as AI becomes more embedded in day-to-day operations, and pressure on ROI increases, the idea of tying spend to measurable performance will only become more intuitive.

For insurers, the real opportunity will lie in strengthening the data, governance and AI foundations that make OaaS possible and then choosing the right areas to apply it. Those that delay this shift risk locking themselves into another decade of high-cost tools with low-impact results.

The insurers that get that balance right will be the ones who turn AI from promise into performance.