Early Stage Investments 

Much pioneering work with generative AI involves experimentation without clearly defined products or tangible deliverables readily meeting capitalization criteria. During these early R&D phases, costs may be more appropriately expensed rather than capitalized as assets. 

Uncertainty in Useful Life   

Useful life estimates for generative AI solutions prove extremely difficult to reliably quantify given the brisk pace of advancement in this space. Whereas conventional policy admin systems typically follow 5+ year depreciation models, AI platforms and their rapidly evolving capabilities often render existing features and functionality obsolete within months. To account for this uncertainty, insurers may find it beneficial to review useful life assumptions annually. 

As Alok Bhargava, a partner at EY commented: “ For Insurance organization to evaluate the useful life of generative AI solutions can be a complex task due to the swift pace of development in this field. Today’s use case, built around a single model, might require the integration of multi-model in immediate future owing to rapid technological advancements. Given the heightened uncertainty regarding the lifespan of these AI solutions, it’s advisable for Insurance carriers to expense, instead of capitalizing their investments in generative AI.” 

For example, in just the year since the introduction of ChatGPT 3.5, numerous enhancements have been added that could obsolete solutions built around 3.5’s original capabilities. Alternatively, future solutions might apply several LLMs working in concert.  

Ongoing Commitments  

Insurance CFOs should also note that relative to conventional on-premise systems, generative AI solutions often require considerably more ongoing investments in maintenance and continuous training. 

In summary, while Datos Insights sees extraordinary value potential in generative AI, we urge insurers to take a prudent financial planning approach when allocating resources to these transformative and rapidly evolving technologies. Please reach out for additional perspectives on AI investment strategies. 

The post The Challenges of Capitalizing Generative AI Investments appeared first on Datos Insights.

Yes, AI will transform the insurance sector. But the biggest disruption will come from how people leverage AI to become far more productive. The insurance professionals who embrace AI as a tool, especially generative AI, will thrive – even without waiting for their company to act. 

Generative AI refers to models that can produce novel content, predictions, and recommendations. Unlike narrow AI designed for specific tasks, generative models can apply learning across domains. Leading generative AI techniques include image, video, text and data generation. 

AI alone still cannot replace most jobs. While generative models are rapidly advancing, general artificial intelligence comparable to flexible human-level reasoning does not exist yet. We remain years away from AI that can fully replicate human cognition and consciousness.    

However, generative AI is increasingly amplifying human capabilities. Rather than directly replacing jobs, AI is empowering people to work faster with higher quality. For example, generative AI tools can help developers draft requirements, code, test and document their applications. Generative writing assistants assist underwriters in summarizing complex submissions and assist agents with marketing copy. Numerous use cases exist though many are beyond the reach of a casual user and require greater expertise and investment. 

Companies investing first in providing tools and upskilling for AI will have a competitive advantage. However, proactive individuals getting a head start with AI can outpace their employer’s training. People proficient in generative AI can become power users compared to peers only comfortable with traditional software.   

You don’t need management approval to augment your unique strengths with AI. Start exploring generative tools offering free trials and take online courses to guide your upskilling. As you become adept at using AI to work smarter, you’ll accomplish more than your job description requires and stand out from the competition. 

So while AI itself does not pose an immediate threat to take jobs at scale, people applying AI creatively surely will displace the skills of individuals who refuse to adapt. Your job security depends less on advances in artificial intelligence and almost entirely on your willingness to make AI part of your personal toolkit.

The post You Won’t Lose Your Job to AI. You’ll Lose Your Job to Someone Using AI appeared first on Datos Insights.

This could immensely accelerate insurance IT projects and lower application development costs. However, while the underlying deep learning advances make this feasible soon, businesses may need to temper expectations around maturity and manage risks.

Key Advances Building Towards This Capability

In the last few years, neural networks have achieved impressive results in analyzing images and video, understanding natural language, logical reasoning, and generating code. Combining these strengths, AI agents could potentially map UI flows and data logic captured from application demos onto the underlying codebase.

Vision systems can now identify on-screen components, actions, and responses. Language models allow translating text and speech into structured knowledge. Reasoning algorithms can map between conceptual representations. And code generation techniques have progressed from simple programs to complex architectures.

Assembling these capabilities into an integrated solution is the next step several AI teams are eyeing. The payoff for automated coding at scale is substantial, evidenced by heightened VC interest in startups like Anthropic working on this technology.

Emergence of New Competition

However, the same capability allowing rapid application development from demonstrative videos could significantly disrupt established technology vendors serving the insurance industry. As barriers to entry into solution creation lower, new players leveraging AI coding could launch competitive alternatives faster and cheaper.

Incumbents may suddenly find their extensive software IP and armies of developers losing differential value. Industry expertise encoded into proprietary systems over years risks being commoditized by AI-fueled startups. Emerging “no-code” tools empowering non-programmers present an existential crisis for current market leaders.

Mitigating Strategies for Protection

Legacy technology companies can adopt defensive innovation strategies to protect market position. Investing in internal AI coding labs could help to compete with external offerings. Acquiring promising startups early in capability development further aids maintaining competitive parity.

Strategic partnerships with AI platform companies, academic labs, and integrators also help monitor, evaluate and ultimately deploy automated coding alongside human developers. Proactively evangelizing tools, use cases and integration blueprints engenders ecosystem support.

Carefully navigated, AI-based software automation could expand markets for new solutions while responsibly balancing job impacts across the insurance technology value chain.

The Path Ahead

Neither hype nor despair is currently warranted when evaluating automatic coding from app demos. What insurance technologists should track is how incremental breakthroughs in causal visual reasoning and software synthesis stack up year-on-year. 

When machine learning models reliably demonstrate the ability to understand complex logic from limited samples, such as a video of an application screen flow, the software development process will be transformed providing an order of magnitude improvement to software development productivity. Continued research paper releases and technology validation experiments will illuminate how close this future actually is.

More advances are required before the core systems vendor community is disrupted and the best option for most insurers remains a vended solution. If you’d like to discuss your policy admin, claims or billing core system needs, please contact me or Mitch Wein to continue the conversation.

The post Core Systems Vendors Beware: AI-Enabled Code Generation Is on the Way appeared first on Datos Insights.

Key Aspects and Impact

The draft regulations seek to limit the use of automated systems, particularly for decisions that carry legal or similar significance for consumers, such as access to insurance. Insurers would have to provide detailed transparency to consumers on the logic and key parameters behind algorithms and get consent. The draft also proposes extensive opt-out and access rights that could enable consumers to object to or get information on algorithmic decisions. 

If finalized in their current form, these rules would require most insurers to overhaul existing practices on using automated decision systems. They would also need to invest more resources in compliance activities around transparency, access management, and providing consumer rights controls. Leaving algorithms unchecked raises fairness issues, but more restrictions also risk limiting innovation and efficiency.

Steps for Insurance Providers

Insurers have time to assess these regulations and prepare for additional oversight of automated decisions. Here are some proactive steps carriers can take:

• Review existing automated decision systems and document usage, logic, data inputs, and role in decisions. Assess validity, fairness, and process rigor. 
• Enhance transparency for consumers on your use of algorithms and develop a compliance strategy for meeting notice, access, and opt-out requirements. 
• Evaluate restricting automated decisions for higher-risk processing like underwriting eligibility until the regulatory environment is clearer.
• Engage industry groups in shaping final rules to find the right balance between protecting consumers and enabling responsible innovation.

Regulatory oversight of algorithms is clearly increasing. Taking early action to align practices with emerging expectations will put insurers in a stronger compliance position. Monitoring developments and participating in policy discussions can also help secure favorable outcomes.

If you’d like to discuss our recommendations for staying ahead of regulatory requirements for AI, please read our report Explaining AI and ML Algorithm Outcomes to Insurance Regulators: CIO/CTO Checklist, or contact me or Mitch Wein to continue the conversation.

The post New Draft Regulations Place Restrictions on Insurers’ Use of AI  appeared first on Datos Insights.

The post Navigating the Evolving Landscape of MGA Core Systems appeared first on Datos Insights.

I recently wrote about business metrics in Aite-Novarica Group’s report Insurer Productivity Metrics, which focuses on four functions—agency management, underwriting, claims, and contact centers—from the perspective of IT involvement.

Internally focused IT metrics are not effective in conveying the productivity value of a technology organization. IT metrics have a place and can enable IT management to measure the performance of their teams to drive efficiencies, improve quality, and establish and monitor service-level agreements—which often have KPIs incorporated into them. From a business perspective, though, what purpose does it serve to tout a low defect rate if the technology does not support improved business outcomes? Or that a developer produced a hundred lines of code when five lines would have sufficed?

While business-oriented metrics can’t solely be measured by IT’s performance, there are some best practices when determining how best to measure the impact of IT. Depending on the metric, consider varying periods such as daily, weekly, or monthly for metrics that are measuring activity. For metrics measuring quality, impact, and activity, consider periods of rolling three months, 12 months, year to date, quarter to date, month to date, or year over year. Many of these metrics can be calculated for individuals, teams, departments, and the enterprise overall.

The definition of “success” can also vary depending on the market, channel, line of business, premium amount, and loss amount. For example, it would be inappropriate to judge rural- and urban-based marketers against the same agency visit count goal. The additional “windshield time” the rural-based business developer incurred would imply a lower target. Perhaps the best measure of success is consistent improvement over time.

As IT becomes a more critical part of delivering insurance services, understanding the performance of those services is the best way to demonstrate that value. The biggest challenge for insurer CIOs is securing budgets, and the key to doing that is by having management and peers acknowledge and recognize the value IT brings. To learn how to apply key, insurance-specific productivity metrics that demonstrate IT’s business impact, access the full report here.

The post Using Productivity Metrics to Measure Insurer IT Success appeared first on Datos Insights.

Largely unburdened by corporate bureaucracies and legacy technologies, MGAs are typically more agile than most insurers, enabling them to address unmet market needs. Many of the product innovations developed by MGAs represent a “c”-change in the rapidly changing risk landscape. In addition to dealing with climate, cyber, cannabis, and crypto, MGAs are in the vanguard of addressing the risk management needs that new business models, such as the sharing and gig economies, are creating.

Combining this nimbleness with an ability to attract strong underwriting talent enables MGAs to move quickly and gain first-mover advantages. There is a class of talented and experienced underwriters attracted to MGAs’ agility and willingness to innovate. And given the significant volume of M&A activity in this space, the opportunity for a significant liquidity event looms large—further enticement to attract the best industry talent.

How MGAs Differ From Traditional Insurers

Many MGAs have coupled their product focus with cloud-based technologies to deploy fit-for-purpose technology solutions. Insurers often rely on legacy technology, cannot integrate via APIs, and are burdened by technical debt. MGAs can introduce new products and product modifications in a fraction of the time it takes traditional markets. The speed-to-market edge significantly contributes to MGAs’ lead with insurance product innovations. Cloud-based platforms eliminate the need for data centers, and Software-as-a-Service pricing models lower up-front capital requirements.

However, MGAs are compelled to be more frugal than traditional insurers as their economics more closely resemble brokers and agents. The ability to deploy highly configurable policy administration systems within the financial constraints of a distributor remains a challenge. The larger, deeper-pocketed market participants want to refresh their technology to maintain relevance and bridge capability gaps.

Like the rest of the industry, MGAs consider robust data and analytics capabilities to be table stakes. Not only does data help produce better underwriting results, but mature data capabilities can make an MGA a more attractive distribution partner for (re)insurers. Program administrators have focused niches and access to a critical mass of data that enables unique insights. The ability to demonstrate with data the quality of the portfolio to (re)insurance partners is highly valued. The availability of cloud-based platforms along with third-party data and analytics are contributing to these mature data capabilities.

Growth Ahead for MGAs

Despite the current economic headwinds, there remains significant interest in the MGA sector. Insurance has traditionally been viewed as an economic sector relatively resilient to recessions and inflation, yet the pressure for profitable growth remains. (Re)insurers and private equity are looking to put capital to work and grow the top line to support the expansion of the channel. Gain sharing creates win-win outcomes for MGAs and capital providers alike. The most attractive MGA partners will be able to demonstrate a track record and will have a seasoned portfolio of profitable business.

Talented underwriters are recognizing the promise of MGAs. They seek an entrepreneurial experience, expanding (re)insurers and investors. MGAs are a source of industry innovation, creative partners, and formidable competitors. 2023 may be remembered as the year of the MGA.

To read more about this and other top trends in property/casualty insurance this year, check out our report Top 10 Trends in Property & Casualty, 2023: Turning Disruption Into Opportunities.

The post MGAs Get All the Attention appeared first on Datos Insights.

Aite-Novarica Group’s new report, Establishing and Sustaining Data Mastery, introduces the Insurance Data & Analytics Maturity Model (iDAMM).

Summary of the Insurance Data & Analytics Maturity Model

This maturity model enables insurers to review where they currently are in their data and analytics journey, define a target state, and plan investments and initiatives to bridge the gaps between them. iDAMM allows insurers to assess their data organization and capabilities across seven dimensions and 21 subdimensions, including leadership and organization, data governance, and architecture and technology management.

The model uses three stages of maturity: Traditional, Evolving, and Transforming. Insurers are likely to have different levels of maturity across different model elements. Moving into a more mature stage is a function of organizational and technological capability, not duration. Like transformations in other parts of the insurance carrier, data and analytics innovation requires enabling technology, organizational change, and executive sponsorship.

Symptoms of Low Data Maturity

Most insurers are falling short of their data objectives despite increasing their focus and investments in this domain.

Organizations that suffer from data immaturity experience symptoms such as data access challenges, low business intelligence and analytics productivity, and spending significant effort to reconcile reports. These organizations may also be noncompliant with data privacy regulations due to ignorance of applicable regulations, incomplete data governance, or gaps in data access controls. Data-immature organizations also suffer from consistent data quality issues and fail to receive the full value of data science efforts and third-party data insights.

The Importance of Culture

Sustaining data mastery requires the insurer to establish and maintain a data culture. Data cultures are those in which data is fully democratized, in which data literacy is high, and that allow all tactical and strategic decision-making to be largely data-driven (but still informed by intuition). Further, all users value data highly and act as corporate data stewards in maintaining high data quality and protection.

Building a data culture is challenging. Insurers that have achieved data mastery will have a culture grounded in data and analytics that can survive leadership changes and shifts in business focus. This means that everyone cares about data quality, as everyone understands innately the value of information and insights. Strategic and tactical decisions are heavily influenced by signals in the data, though intuition developed through industry experience also plays a role.

For more information on data mastery and the iDAMM, access the full report here.

The post Introducing the Insurance Data & Analytics Maturity Model appeared first on Datos Insights.

Data KPI Snap Poll Methodology

As organizations embark on data transformations or develop enterprise-wide data strategies, it’s important to understand how your organization currently uses data to determine where gaps exist and how you can get the most business impact. I reached out to 25 data leaders at property/casualty insurers to learn about their data practices and key performance indicators (KPIs). Of those contacted, five responded to share insights about the data assets they currently have in production, how they track asset utilization, how they measure data quality, and what other KPIs they use to measure success.

Production Data Assets

All respondents listed operational reports, dashboards, data warehouses, and data lakes as data assets currently in production. The majority listed self-service business intelligence (BI) and data marts as being data assets in production. Few listed actuarial sandbox and data catalog, and none listed data lakehouse.

Data Utilization KPIs

When I inquired about tracking, most respondents stated that they do implement tracking in various ways. One respondent mentioned that their tracking mechanisms are not yet fully implemented but shared their plans to focus on utilization statistics such as query volume, user numbers, and report view totals. Other respondents identified a number of ways to track data asset utilization and adoption, including the number of registered users accessing repetitive assets, system-based usage reports, what computer resources are used, and the creation of data governance councils.

Measuring Data Quality

There was less measurement of data quality among respondents, but all acknowledged that it was an important element of the overall data strategy and expressed plans to implement metrics in the future. One respondent did list financial data as currently being highly governed and monitored but stated that there was far less monitoring of other data domains. Another respondent stated that their organization’s data quality focus centered on client data, with metrics for accuracy, completeness, and validity.

Determining the Value of Data

Two respondents stated that while having formal KPIs to measure impact is a strategic goal, they had not yet defined or implemented such metrics. One of our respondents noted that program success is measured by its impact on other business KPIs, but they stated that there was still too much fluidity to their metrics. Their goal is to move quickly toward using quantifiable impact or return on investment (ROI) to drive organizational decision-making. Additionally, one respondent also measured business value from specific use cases to determine impact.

Concluding Thoughts

As strategic data usage becomes more embedded in insurers’ daily operations, developing a clear set of KPIs to track business value will become increasingly important. To learn more about our research around data and its impact, read our recent report Establishing and Sustaining Data Mastery: Introducing the Insurance Data & Analytics Maturity Model.

The post Insurance Data Leaders Are Implementing Data KPIs appeared first on Datos Insights.

During this meeting, we introduced an upcoming Aite-Novarica report, the Data and Analytics Maturity Model, and also had a discussion about data lakehouses.

Introducing the Data and Analytics Maturity Model

The Data and Analytics Maturity Model will follow the format of previous Aite-Novarica maturity model reports (see, for example, the Agile Maturity Model for Insurers, January 2020). Data maturity is assessed across seven dimensions and 21 subdimensions, and stages of maturity are divided into Traditional, Evolving, and Transforming.

Data maturity is technology agnostic. It addresses technology selection and modernization, processes, and human capital. It’s not just about having a data lakehouse—we’ve seen organizations with the latest and greatest technology still not meeting the needs of their business users. And conversely, we’ve encountered organizations with traditional data environments that excel at supporting evolving business needs. The key to achieving data maturity is to consider all these capabilities and balance them. Organizations don’t need to be Transforming in all capabilities—they should align their transformation with the business strategy.

One of our Special Interest Group attendees provided the following feedback on the model: “All the components of having strong data analytics capabilities are represented. It shows us how we are making progress over time and gives us something to bring back to our partners and C-suite to measure that across time.”

We look forward to sharing the full Data and Analytics Maturity Model report soon.

The Data Lakehouse

Our discussion then moved on to a discussion of the data lakehouse. Recent advances in technology, driven by cloud hosting and cloud economics, have enabled the data lakehouse to emerge as a practical solution to meeting insurers’ data and analytics needs.

A poll of the attendees at our Special Interest Group found that 42% of attendees’ organizations have a cloud-based lake, 17% have an on-premises lake, and 42% do not have a lake.

At Aite-Novarica, we are not seeing any new data lake developments on-premises. There is high penetration of cloud-based data lakes—and more recently lakehouses. Almost two-thirds of insurers have data lakes at their disposal, and we are seeing great investment in this area. For more information on the data lakehouse, see our Aite-Novarica Group report, The Data Lakehouse: How Past Data Architectures Produced a New Paradigm (July, 2022).

Concluding Thoughts

Data and analytics will become even more important to insurers and be key to competitive differentiation and advantage. The industry is moving very quickly, and it’s important for insurers to stay abreast of what is going on out there so they don’t fall behind, which requires leadership, funding, and skills. To ensure success, investments in data and analytics must be driven by business need and value.

For more information about future Aite-Novarica events, visit our Events page.

The post Highlights From the Insurance Data and Analytics Special Interest Group appeared first on Datos Insights.

Emulating strategies employed by enterprise software solution providers such as IBM and Oracle, Broadcom’s expansion of its software portfolio enables further cross-selling and margin expansion of products that corporate IT departments find challenging to swap out. Once a vendor has a significant presence in an organization’s data center or technology stack, the vendor can leverage that to compel the purchase of additional products or force significant price increases at licensing renewal.

It is commonplace for acquisitions to trigger a new round of license audits as acquirers are eager to better understand their client base and accelerate value realization. Insurers that have failed to manage their software license may be at risk of adverse licensing audit results. Non-compliance provides insurers with a Faustian choice: pay the penalties or buy additional licenses.

Technology audits are difficult to manage as each vendor measures “usage” differently; they may also have multiple metrics for each product. A midsized IT organization may be dealing with over 100 different licensing metrics. Furthermore, certain usage metrics may be difficult to measure, and the software license monitoring solution provided by the vendor may be sensitive to improper installations.

A number of tactics can be taken to optimize audit outcomes. These include:

• Proactive management of the auditor
• Building a cross-functional team to understand entitlements and usage
• Scrutinizing audit results
• Proposing win-win alternatives to the software vendor

Technology vendors have every right to be compensated for the usage of their products. Given technology’s nature, product usage can proliferate within an organization without the knowledge of the vendor or the carrier’s IT or procurement organizations. Audits are a means to capture this information and charge the carrier appropriately.

Broadcom’s VMware acquisition is likely to trigger a wave of licensing audits. These audits are disruptive, and they consume time and money. Taking proactive action to minimize these events is appropriate. Insurer IT organizations who have scrutinized their auditors’ initial findings have been able to reduce the ultimate cost by as much as 90%. As they say, “Knowledge is power,” and this may be truer in technology licensing audits than anywhere else.

For further suggestions on how to best handle a software licensing audit, please read Aite-Novarica’s report CIO Checklist: Software Licensing Audits or reach out to me at eweisburg@datos-insights.com.

The post What Insurers Should Take From the Broadcom-VMware Acquisition appeared first on Datos Insights.