Info-Tech guide tackles fragmented data strategy for P&C insurers

Guidewire puts the drag at up to 80% of a data practitioner’s time spent preparing and transforming data, leaving only 20% for strategic analysis. Fragmented data is still the hidden brake on P&C modernization, and Info-Tech Research Group’s Build a Data Strategy for Property & Casualty Insurance is built around that problem. Data strategy is a business discipline, not a technical side project, because carriers cannot trust AI, reporting, or underwriting decisions if policy, claims, billing, underwriting, customer, and third-party data remain disconnected.

Why fragmented data keeps slowing P&C software programs

The basic failure mode is familiar across P&C cores: each function accumulates its own definitions, ownership, and quality rules, then everyone spends time reconciling the differences later. That drag has made data platform and analytics products part of the core insurance software conversation, not a separate warehouse discussion.

The Casualty Actuarial Society addressed the same issue years ago in Monograph Series Number 9, Data Quality Management in the P&C Insurance Sector. If the insurer cannot govern the data at the source, every downstream system, from pricing models to claims dashboards, inherits the same ambiguity.

The domains a usable strategy has to connect

In P&C, the domains that matter are policy, claims, billing, underwriting, customer, and third-party data, and a usable strategy has to show how each one is owned, validated, and linked. That matters because these records are rarely aligned cleanly across core systems, especially when a carrier has inherited multiple legacy platforms through years of product and organizational layering.

A usable strategy also has to answer a set of practical questions at the same time:

• Which data domains are business-critical
• Who owns each domain and approves changes
• What quality standards apply across systems
• How lineage and metadata will stay visible for audit and compliance
• Which foundations have to be in place before AI, automation, or advanced analytics can scale

A usable strategy does not stop at describing governance in abstract terms; it requires architecture, stewardship, and operating-model decisions that turn fragmented records into decision-grade information.

Why governance moved into the buying conversation

Governance is no longer treated as an afterthought in P&C software buying. In a June 6, 2026 article, Perceptive Analytics called data governance a top-three buying criterion for P&C insurance CIOs, which matches what carriers are experiencing as they evaluate core replacements, analytics layers, and AI tools. Once the buyer is accountable for lineage, retention, and definitions, governance becomes a procurement issue as much as an IT design issue.

Lineage, metadata, and compliance now run through insurance data materials. Atlan and Guidewire both treat those capabilities as practical requirements, not optional architecture decorations. If an insurer cannot trace where a field came from, who changed it, or which system is authoritative, then it cannot reliably defend a report, explain a model, or control operational risk.

AI, underwriting, and claims all depend on the same foundation

The 2025 and 2026 pressure on carriers has been to use AI more effectively, but the data problem has not gone away. Better underwriting models, faster claims triage, and more personalized customer engagement all depend on data definitions that mean the same thing across systems. When the policy system, claims system, and billing system each describe the customer or the exposure differently, AI simply automates the inconsistency.

That governance pressure is reinforced by regulation. On March 18, 2025, Wisconsin Commissioner of Insurance Nathan Houdek issued Bulletin 20250318AI, The Use of Artificial Intelligence Systems in Insurance. If an insurer wants to deploy AI in underwriting or claims, it needs clean ownership, traceable lineage, and documented control over how data feeds the model.

How modernization work turns strategy into operating discipline

The stronger programs tie governance to architecture and then to measurable outcomes such as cleaner regulatory reports, better financial controls, and faster integration of new applications. In a core replacement, that can determine whether a new platform becomes a single source of truth or just another interface on top of the same old fragmentation.

Intellias’ March 26, 2025 piece on data modernization in insurance framed upgraded data management systems as part of future-readiness, and Milliman’s March 20, 2026 article on data quality in the insurance sector argued that machine learning and AI can help improve it.