PERCEPTIVE ANALYTICS PERSPECTIVE The Bottleneck Is Not Underwriting. It’s the Data. At Perceptive Analytics, we work with mid-market P&C carriers who are ambitious about growth but are hitting data infrastructure limitations Across multiple engagements, we have seen that underwriting speed stalls because the incoming data is messy, incomplete, and locked in static formats. Your underwriters have the expertise to make decisions, but they cannot use it quickly when they have to dig through unstructured files. A slow quote is just the symptom. The real issue is your unstructured submission intake process. The carriers winning market share today do not win because they have better underwriting expertise. They win because they treat submission data as a clean, structured asset. They build their workflows on top of that clear data. This guide explains why you need to make this transition now, what it takes to get there, and how to plan your investment. |
1. The Mid-Market Growth Paradox: More Submissions, Slower Decisions
There is a common scenario that happens across multiple growing mid-market P&C carriers that poses a major challenge at the executive level. Premium volume and new business pipelines are expanding. Broker relationships are active. Specialty and regional commercial lines are drawing new submission flows. Yet, their underwriting throughput (the number of risks your team can actually evaluate, price, and bind) is falling behind. In many carriers, it is actively deteriorating.
This problem does not stem from a lack of talent or appetite. It comes from process architecture. Submission volumes have shot up, but underwriting teams have not grown to match them. According to the Global Insurance Report 2023 from McKinsey, commercial P&C premiums grew by 6% to 8% annually starting in 2018 [McKinsey Report] . Other industry reports suggest that MGA premium volumes in the US grew by roughly 14% annually over the last ten years. Despite these massive volumes, the workflows you use to process these files remain unchanged [Reinsurance]
What Is Driving Submission Growth
Brokers are shopping in more markets than before. This is especially true for commercial property, professional liability, and specialty risks, where coverage has tightened. The Council of Insurance Agents and Brokers (CIAB) Q1 2024 Market Survey confirmed that commercial property underwriters now require far more documentation than they did in previous cycles. Brokers are quote-shopping aggressively. At the same time, broker mergers have created massive distribution firms that route the same submission to dozens of carriers at once.
The Headcount Constraint Is Not Going Away
Hiring more underwriters is rarely an option. Experienced commercial lines underwriters are hard to find and expensive to hire. The talent pipeline has shrunk over the last decade, and retiring seniors are leaving without passing down their knowledge. You cannot solve this by adding more salaries to match submission growth. Doing so would ruin your expense ratios and eat into your margins.
40% of underwriters’ time spent on non-core and administrative tasks — not risk assessment Accenture, 2022 [1] | 6–8% annual commercial P&C premium growth since 2018, compressing underwriting capacity McKinsey, 2023 [4] |
The strategic implication is direct: growth in submission volume is not translating into scalable underwriting throughput. The bottleneck lies in the manual process sitting between the broker’s email and your underwriter’s desk.
2. The Hidden Bottleneck: Submission Data Chaos
If you walk through the intake process at a typical mid-market carrier, you will see a process that has not changed in fifteen years. A broker sends an email. It contains a mix of ACORD PDFs, loss run spreadsheets, supplemental questionnaires, and free-form text descriptions. None of these documents use a standard format. None of them can be read by a computer. A human has to read, interpret, and sort every single file before any of your internal systems can use the information.
This is a structural problem built into the commercial insurance market. Most carriers build their intake processes to tolerate this chaos instead of fixing it. This decision gets more expensive with every new email that hits the inbox.
The Anatomy of a Fragmented Intake
A typical commercial lines submission packet for a mid-market account might include: a broker cover email with contextual notes, a completed ACORD 125 (Commercial Insurance Application) in PDF form, supplemental forms specific to the line of business, three to five years of loss runs from the incumbent carrier, schedule of values or statement of values for property risks, and increasingly environmental or operational data the broker has assembled from the insured directly.
None of this arrives organized or ready for your software. This causes several operational headaches:
- Manual triage: Your staff must read each submission to identify the line of business, class of risk, and destination.
- Manual data entry: Your team must copy risk data by hand into your policy administration or rating systems, which wastes time and leads to typing errors.
- Chasing missing data: Submissions are often incomplete. Your team has to email the broker, wait for a response, and put the file back in the queue, adding days to your turnaround time.
- Inconsistent risk data: When different people enter data at varying levels of detail, your risk profiles become inconsistent. This makes portfolio analytics and reporting inaccurate.
- Delayed routing: Without automated classification, submissions sit in shared inboxes until someone manually forwards them to the right team.
The issue is not the submission volume itself. The problem is the inconsistency of the incoming files. A carrier receiving 500 clean, system-ready submissions faces a simple task, while a carrier receiving 200 fragmented, multi-document emails faces a labor-intensive chore.
PERCEPTIVE ANALYTICS PERSPECTIVE Inconsistency Is the Real Expense and Few Carriers Are Measuring It Correctly Most carriers account for underwriting intake costs as a headcount line. That framing obscures the true cost of submission variability. When an underwriting assistant spends forty minutes cleaning up a messy submission, chasing a broker for details, typing ACORD fields, and double-checking loss runs, you label that time “underwriting operations.” It never appears on your balance sheet as the “cost of bad data.” At Perceptive Analytics, we advise our clients to calculate the processing cost per account by separating clean submissions from those that require manual clean-up. In our experience, messy submissions cost two to four times more to process than structured ones. This cost accrues before an underwriter even evaluates the risk. This is the metric your executive team needs to watch. |
3. Why Legacy Intake Workflows Break at Scale
When submission processing becomes visibly painful, carriers respond. The interventions are predictable and well-intentioned: hire additional underwriting assistants, invest in a document management tool, expand the shared services centre, or deploy rule-based workflow routing. These are not wrong decisions in isolation. While these actions make sense in the short term, they act as temporary patches. They hide the core data issue rather than solving it.
Adding Headcount: Optimising Around the Problem
Hiring underwriting assistants buys capacity, not capability. The manual intake process consumes more staff time as volume grows, but the core variability problem remains. Each new hire must be trained to interpret and normalise fragmented submissions, these are skills that take time to develop and are not easily transferred. When a senior underwriting assistant leaves, institutional knowledge leaves with them. The answer to a data problem is not more humans processing bad data.
Rule-Based Workflow Automation: Brittle by Design
Rules-based routing fails when a submission deviates from the expected format. In commercial lines, exception rates for rules-based tools regularly hit 30% to 50% of total volume. Every single exception gets kicked back to a human for manual handling. As you expand into new territories or specialty lines, your exception rate climbs. The system designed to save time ends up creating a more complex manual review queue.
Shared Service Expansion: Scaling Inefficiency
Consolidating intake into a shared service centre can create apparent efficiency through volume aggregation. But shared service models built on manual intake still require trained staff to interpret unstructured submissions, still generate re-keying errors, and still introduce processing latency as submissions queue between intake and routing. The Capgemini Insurance Top Trends 2024 report identifies data integration and workflow automation as the two most underdeveloped capabilities in commercial lines operations [5], Most still rely on human eyes to classify incoming documents, regardless of how large their service centers are.
The conclusion, for any carrier operating at growth, is uncomfortable but clear: you cannot automate what is fundamentally inconsistent and unstructured. The path forward requires fixing the data, not adding layers around it.
4. Structured Submission Data: The Foundation of Scalable Underwriting
Structured submission data is information extracted from emails, PDFs, forms, loss runs, and questionnaires, then organized into a clean, system-readable format. This structured output is validated for accuracy and fed directly into your policy systems, rating tools, and analytics engines.
The format of your incoming data matters. A PDF copy of an ACORD form is not structured data. The moment an employee has to read that PDF and type the details into your system, you are running a document-entry business instead of an insurance business. Structured data removes that manual typing step, transforming your operational economics.
What Structured Submission Data Enables
- Automated classification: Your systems identify the risk type, business line, and account size immediately at ingestion.
- Instant completeness checks: Your software flags missing data and emails the broker for updates automatically, saving your staff from doing it.
- Smarter routing: Your systems score and prioritize incoming submissions based on your underwriting appetite, sending them to the right desk without manual triage.
- Instant data enrichment: Your system matches structured data to third-party services like Verisk, ISO, or geospatial models automatically, sparing underwriters from manual web searches.
- Straight-through processing: For simple, low-hazard risks, your systems can generate quotes automatically, allowing your underwriters to focus on complex accounts.
The Submission Pipeline: From Document to Decision
The transformation can be visualised as a structured pipeline that converts raw submission packets into decision-ready assets:
| Submission Packet | → | Data Extraction | → | Standardisation & Validation | → | Intelligent Routing | → | Underwriting Decision |
Each stage in this pipeline is a place where value is either created or lost. Without structure at the extraction and standardisation stages, every downstream step — validation, routing, enrichment, and the underwriting decision itself — carries the cost of human compensation for data gaps.
5. What Structured Submission Data Unlocks for Mid-Market Carriers
Moving to structured data is an operational upgrade that delivers clear, financial returns. We see five major impacts for mid-market carriers:
- a) Faster Quote Turnaround
Intake processing lag is the single largest contributor to quote cycle time for most commercial lines submissions. When risk data arrives pre-structured, the time between submission receipt and underwriter engagement compresses significantly. Industry practitioner benchmarks suggest that structured intake can reduce intake processing time translating directly into faster first-touch underwriting and, ultimately, faster quote delivery. This is also something that Brokers notice. Ivans’ 2025 Insurance Agency-Carrier Connectivity Trends Survey found that real-time appetite information and faster commercial submission processing are now the top two factors agents cite when selecting carrier partners [6].
- b) Improved Underwriter Productivity
Accenture reports that underwriters spend 40% of their day on administrative work, a metric that has not changed since 2008. Structured data eliminates this clerical burden. When your underwriters receive files that are already extracted, validated, and enriched, they can spend their time analyzing risk instead of typing data. Capgemini indicates that automation can eliminate up to 43% of an underwriter’s administrative workload.
70% of underwriters’ total work time spent on non-underwriting activities — administration, negotiation and sales support combined Accenture / The Institutes Longitudinal Underwriting Survey, 2024 |
- c) Better Broker Experience
Speed is your best asset when competing for broker business. If your system processes a submission and returns a quote or declination within hours, brokers will send you their best accounts first. The Ivans survey shows that 72% of agencies want to see more commercial submission automation from their carriers. If your business falls behind on speed, brokers will route their best risks to faster competitors, leaving you with lower-quality accounts.
- d) Greater Portfolio Consistency
When your risk data flows through a clean, automated pipeline, your underwriting files remain consistent across your business. Your portfolio analytics, catastrophe modeling, and reinsurance reporting become far more accurate. Inconsistent data inputs lead to unreliable analytics. If you are managing reinsurance limits or regulatory reports, your business decisions are only as good as the raw data you feed into your models.
- e) Operational Scalability Without Linear Headcount Growth
This is where the financial benefits become clear. Structured submission data allows you to scale up your incoming business without hiring more operations staff. With clean data, straight-through processing rates for simple commercial accounts can reach 30% or higher. Your staff can then focus on complex risks where human judgment is actually required. Cutting your intake processing times by 30% to 50% across your entire portfolio will improve your expense ratios. McKinsey estimates that automation in underwriting can save the global insurance industry $9 billion to $15 billion annually, with the biggest benefits going to early adopters.
6. Why This Matters More for Mid-Market Carriers Than Tier-1 Insurers
It is tempting to frame structured submission data as a capability that large carriers will build first and mid-market carriers will adopt later, once the path is proven. That framing misreads the competitive dynamics. The urgency is, in important respects, higher for mid-market carriers.
Leaner Teams, Less Operational Buffer
Tier-1 carriers operate with underwriting teams large enough to absorb process friction through sheer headcount. They can staff around inefficient intake, fund specialist roles for submission triage, and build shared service infrastructure that partially compensates for unstructured data. A mid-market carrier with fifteen or twenty commercial lines underwriters does not have that buffer. When intake absorbs forty percent of underwriter time, the remaining capacity is simply insufficient to handle growing submission volumes without deteriorating service levels.
Greater Margin Sensitivity
Combined ratios at mid-market carriers leave less room for expense ratio drift than at carriers with scale-driven expense advantages. Process inefficiency is not just an operational inconvenience it is also a margin leak. NAIC data confirms that commercial lines expense ratios at smaller and mid-market carriers consistently run higher than those at top-tier national carriers [10], making operational efficiency an existential rather than aspirational priority.
Broker Relationship Risk
Mid-market carriers rely on independent brokers. These relationships depend on speed. Brokers expect you to review their submissions quickly, offer a competitive price, or issue a fast declination. Ivans reports that 60% of agents feel their carrier partners are too difficult to work with. If you lose your fast-response reputation with your top brokers, your incoming pipeline will deteriorate quickly.
PERCEPTIVE ANALYTICS PERSPECTIVE The Competitive Moat Is Narrowing Faster Than Most Executives Realise The carriers that have already invested in submission data infrastructure are not advertising it. They are simply quoting faster, declining earlier, and winning preferred status with brokers who route their best accounts to carriers that respond. The lag between ‘this is a capability advantage’ and ‘this is table stakes’ in commercial insurance technology has historically been five to seven years. That window is compressing. Structured submission data is not a digital transformation luxury for mid-market carriers but it is a competitive necessity. The carriers that treat it as an infrastructure investment today will be quoting at the speed of data, not the speed of manual processing, when the next hard market cycle creates another surge in submission volume. Those that wait will be absorbing that surge with the same fragmented intake architecture they have today. |
7. The Build-vs-Buy Decision: Avoiding the Wrong Technology Bet
Once the strategic case for structured submission data is accepted, the implementation question follows immediately: build internally, buy a point solution, or partner with a domain-specialist provider? Each path carries distinct risks, and choosing the wrong one wastes time and capital that mid-market carriers cannot afford to lose.
The Risks of Building Internally
Internal build programmes appeal to carriers with strong data science teams and a preference for bespoke solutions. The appeal is understandable where one can opt for a proprietary extraction model that can, in theory, be tuned precisely to the carrier’s specific lines, forms, and data requirements. In practice, the delivery risks are significant. Commercial lines submission variability is exceptionally high: ACORD form versions differ, broker-supplied documents vary in structure and quality, and the edge cases in specialty lines can overwhelm a rules-based or lightly-trained model. Building a production-grade extraction and normalisation system for commercial lines data typically requires twelve to twenty-four months of development, a dedicated data science and engineering team, and an ongoing maintenance commitment as submission formats evolve. Most mid-market carriers do not have this resource profile.
The Risks of Point Solution Procurement
The Insurtech market offers numerous point solutions targeting submission intake like OCR tools, document classification APIs, and intake workflow platforms. The procurement risk lies in narrow scope and integration complexity. A tool that accurately extracts data from ACORD 125 forms may fail on supplemental questionnaires, loss runs, or the free-form broker emails that carry critical account context. Integration into legacy policy administration systems, the infrastructure reality at most mid-market carriers frequently requires custom engineering that is not included in the vendor’s scope. Line-of-business fit is another constraint: a solution designed for small commercial may perform poorly on mid-market specialty risks.
What to Look for in a Domain-Specialist Partner
The more defensible path for most mid-market carriers is partnership with a provider that combines domain-specific data engineering with insurance-native extraction models and demonstrable integration flexibility. The key evaluation criteria are: extraction accuracy across the full spectrum of submission document types (not just ACORD forms), configurable output schema that maps to the carrier’s existing systems, coverage of the specific commercial lines in scope, and a deployment model that does not require the carrier to build and maintain the underlying models. Deloitte’s 2024 Global Insurance Outlook identifies AI-enabled intake and data extraction as one of the three technology investment priorities with the highest expected ROI for commercial lines carriers through 2026 [11].
8. A Practical Maturity Roadmap for Structured Submission Transformation
Transformation does not happen in a single implementation. For most mid-market carriers, the realistic path moves through four stages of increasing capability — each building on the data infrastructure established in the prior stage. The following maturity model provides a practical framework for assessing current state and sequencing investment.
| 1 | Manual Intake | Document-centric operations. Submissions arrive via email and PDF. Underwriters triage, re-key, and route manually. High latency, high variability, no visibility into queue status. |
| 2 | Assisted Extraction | Partial automation via OCR and RPA. Data extraction improves speed but rules-based logic breaks on unstructured variance. Human review remains for most submissions. |
| 3 | Intelligent Triage | AI-assisted classification and intelligent routing. Completeness checks automated. Underwriters receive pre-validated, enriched submissions. STP possible on simpler risks. |
| 4 | Decision-Ready Underwriting | Fully structured intake ecosystem. Every submission normalized, validated, enriched, and routed before it touches an underwriter. Portfolio consistency and scalable throughput achieved. |
Sequencing the Journey
If your business is currently at Stage 1, do not try to jump straight to Stage 4. Your first priority should be basic data extraction and standardization: setting up a clean, machine-readable data format that your existing systems can ingest. Moving from Stage 2 to Stage 3 often delivers the fastest financial return. Automated triage cuts down queue times, and automatic completeness checks eliminate back-and-forth emails with brokers.
Moving from Stage 1 to Stage 3 typically takes twelve to eighteen months with the right partner. Reaching Stage 4 depends on the complexity of your specialty lines, but leading carriers are achieving this milestone within thirty-six months of starting their projects.
9. What Forward-Looking Carriers Are Doing Now
The leading edge of the market is not waiting. A defined set of behaviours separates carriers that are structurally positioning for the next growth cycle from those that are managing to the current one.
Investing in Submission Ingestion Intelligence
The most visible investment among leading carriers is in submission ingestion infrastructure — tools and systems that receive, extract, and normalise submission data at the point of entry, before it reaches an underwriting queue. McKinsey’s analysis of AI adoption in commercial insurance identifies faster submission handling, automated documentation, and refined segmentation as the four highest-ROI AI applications in underwriting, all of which depend on structured data as a prerequisite. Carriers that have not yet made this investment are operating a manual step in a process their competitors are automating.
Building AI-Assisted Risk Extraction Capability
Beyond OCR and basic form extraction, leading carriers are investing in AI models capable of interpreting unstructured broker narrative, identifying coverage triggers in loss run descriptions, and surfacing risk characteristics that would otherwise require underwriter reading time.This helps them build a rich library of structured historical data, creating a massive competitive advantage.
Standardising Data Pipelines for Portfolio Analytics
Forward-looking companies know that clean data does more than speed up their daily intake. It acts as the database foundation for their entire analytical operation. Swiss Re’s research highlights data standardization as a top requirement for predictive underwriting. The carriers building these clean pipelines today are setting themselves up to deploy more advanced pricing, risk aggregation, and automated selection models in the future
PERCEPTIVE ANALYTICS PERSPECTIVE The First-Mover Advantage in Submission Data Is Real and Time-Bounded The window of opportunity to build a competitive advantage here is closing. As structured data tools become standard across the industry, the benefit will shift from “we quote faster” to “we have better historical data, better models, and more accurate pricing.” Carriers that build their structured databases early will train better models and secure deeper broker loyalty. Perceptive Analytics works with carriers at every stage of this maturity journey. Our consistent observation is that the carriers who delay because the investment feels large are the same ones who, three years later, are spending significantly more to catch up while managing the operational and competitive costs of continued manual intake.The cost of doing nothing is never zero; it is simply hidden in your expense ratios. |
10. Scale Begins at Intake
For mid-market P&C carriers, underwriting scalability is no longer constrained by underwriting expertise alone. The expertise exists. The appetite exists. What constrains growth is the ability to transform raw, messy submission files into structured, system-ready data before your underwriters ever touch them.
The math is simple. Underwriters who spend 40% of their day on administrative work cannot focus on writing profitable business. Submissions stuck in unread emails are quotes you cannot deliver. Slow responses mean your brokers will route their best risks to your competitors. You can solve every single one of these problems, but not by hiring more staff, installing rigid rules-based routing, or outsourcing your intake. You have to fix your incoming data at the source.
The carriers that invest in data extraction, validation, and automated triage today will operate with a permanent cost and speed advantage as industry volumes continue to grow. Those that continue to manage around manual, unstructured workflows will find that every new submission brings more overhead instead of more profit.
Scale Begins at Intake The carriers that modernize their submission data systems today will lead the market in speed, profitability, and broker satisfaction tomorrow. To discuss how Perceptive Analytics can accelerate your submission data transformation, contact us at perceptive-analytics.com |
Sources & References
All statistics cited in this document have been verified against the original named sources listed below. Data drawn from secondary summaries has been cross-referenced to primary source materials where available.
© 2026 Perceptive Analytics. All rights reserved. This document is provided for informational purposes only and does not constitute legal, regulatory, or investment advice. Statistics are attributed to named third-party sources; Perceptive Analytics makes no independent warranty as to their accuracy. Reproduction in whole or in part requires prior written consent.
Frequently Asked Questions About P&C Carrier Underwriting and Structured Data
1. Why do mid-market P&C carriers struggle to scale underwriting operations?
Mid-market P&C carriers hit scaling walls because their intake process relies on human teams manually interpreting unstructured documents like ACORD PDFs, loss runs, and emails. This manual dependency creates a massive operational backlog, delays quote response times, and damages broker relationships. Perceptive Analytics eliminates this friction by engineering native extraction pipelines that convert unstructured submission chaos into clean, system-ready data assets instantly.
2. What is structured submission data in commercial insurance underwriting?
Structured submission data is raw risk information (exposures, policy limits, schedules of values, and loss histories) extracted from loose PDFs or emails, validated for accuracy, and delivered in standardized, machine-readable formats. Perceptive Analytics builds these automated ingestion pipelines, allowing carriers to route clean data directly into rating engines and core administration software without manual data entry.
3. How does automated submission ingestion improve carrier expense ratios?
Automated ingestion directly reduces expense ratios by removing clerical and administrative tasks, which consume up to 40% of an underwriter’s working day. By automating document classification, extraction, and completeness checks, Perceptive Analytics enables mid-market carriers to scale intake volume and capture market share without linearly growing operational headcount or expanding expensive shared services.
4. Can data extraction layers integrate safely into legacy core policy systems?
Yes, modern automated data extraction ecosystems integrate seamlessly into legacy policy administration systems and modern rating tools via secure data lakes, cloud pipelines, or standard REST APIs. Perceptive Analytics builds custom, non-disruptive integration architectures that feed structured broker risk criteria straight into existing underwriting queues without needing complete, expensive core platform overhauls.
5. What is the primary risk of building a submission extraction system internally?
The primary risk of an internal build program is high system failure rates due to extreme commercial document variability across different broker networks. Building a production-grade extraction engine requires massive, multi-year commitments from data science and engineering teams to handle edge cases. Partnering with Perceptive Analytics mitigates this risk by deploying proven, insurance-native models with flexible output schemas.