Home > Insurance > What to Expect From a Consulting Firm Focused on Combined Ratio Improvement

A decision-stage guide for CFOs, CUOs, and Heads of Transformation at mid-to-large P&C insurers and MGAs.

Perceptive Analytics Point of View
Most combined ratio consulting engagements fail because the diagnostic is shallow. Firms arrive with frameworks built for other industries, rename a few slides, and deliver a strategy deck that the carrier’s own actuaries could have produced. The difference between a 2-point improvement and a 5-point improvement almost always comes down to data granularity and implementation depth — not strategic insight.
You should be able to answer three things before you hire anyone. Where exactly is your leakage by line and distribution channel? Can the firm show they have actually moved the needle during implementation, not just on paper? What skills will your team keep when the consultants leave? If they cannot give you specific answers, you are buying a report — not a result.

You are likely facing board pressure to fix your combined ratio. Your team is busy. Hiring help makes sense if you structure it right. Most carriers end up with either a decent action plan or an 18-month project that moves the needle by 3 to 5 points. The gap between those two outcomes comes down to being clear about scope, cost, and risk from day one.

This guide explains what a useful engagement covers, how to track progress, and how to weigh the cost against your ROI. It is for CFOs and underwriting leaders at P&C carriers and MGAs who need to pick a partner now. Perceptive Analytics works at the data infrastructure layer that makes analytical improvement possible and sustainable — the foundation that separates a plan from a result. You can explore our approach in our data-driven blueprint for growth in the insurance industry and our insurance analytics solutions practice.

96.5% U.S. P&C industry net combined ratio in 2024 — best since 2013 (S&P Global Market Intelligence, 2025)	3–5 pts loss ratio improvement possible through better analytics (McKinsey Global Insurance Report, 2022)	110.1% other liability combined ratio in 2024 — highest since 2016 (S&P Global Market Intelligence, 2025)

Talk with our consultants today. Book a session with our experts now. → Schedule Your Free 30-Minute Session with Perceptive Analytics

1. Core Scope: What a Combined Ratio Consulting Engagement Actually Covers

This is not a strategy project. It is an operational fix. A good program gives you three things: a baseline that shows exactly where you are losing money, a list of tasks ranked by their dollar value, and a plan that pays for itself with early savings. Consultants should find where your ratio is leaking. Most mid-market carriers lose money in two or three specific spots. The goal is to separate bad luck from bad processes.

Diagnostic Scope

A real diagnostic looks at loss and expense ratios separately. On the loss side, you analyze your book by line, geography, and channel to find rate gaps or bad risk selection. On the expense side, you map costs like acquisition and overhead to the actual work driving them. Perceptive Analytics’ advanced analytics consulting team is specifically oriented toward this kind of granular diagnostic work — building the segmented analytical views that reveal where leakage is actually occurring rather than where it appears to be occurring at the aggregate level. Our insurance sales dashboard case study shows what that diagnostic visibility looks like when it is operational.

Underwriting and Pricing

Pricing errors compound over time. The focus here is on closing gaps in how you price risks and ensuring your underwriters are applying consistent standards rather than diverging across teams. One North American carrier used better frontline routines to stop price swings — and that discipline was the only prerequisite for any durable loss ratio improvement downstream. Perceptive Analytics’ AI consulting services support the model development layer that makes pricing consistency scalable rather than dependent on individual underwriter behavior.

Claims and LAE

Claims are usually where the most recoverable money sits. This includes using triage models to route complex claims to senior adjusters early and identifying subrogation opportunities faster. One auto insurer saved nearly $80 million over five years by cutting three days off their total loss cycle [Deloitte, 2023]. Our from reports to real-time: how AI is rewiring the insurance claim process analysis covers exactly how these triage and routing improvements are being operationalized by leading carriers.

Expense and Operations

This covers vendor networks, repair costs, and management layers. You can often cut base spending by 7% to 11% by examining how you handle vendors and internal demand [McKinsey, 2020]. Perceptive Analytics’ Snowflake consulting and Talend consulting capabilities support the data engineering layer that makes expense visibility reliable and continuous rather than dependent on quarterly manual reporting cycles.

2. Methods and Levers: How Consultants Improve Loss and Expense Ratios

The method matters as much as the lever. A firm using interviews and process maps will produce weaker results than a firm running models on your actual claims data. The math matters — and it starts with the quality and granularity of the underlying data infrastructure.

Predictive Modelling and Segmentation

The most effective way to fix underwriting is to model loss ratios by risk segment, building pricing models based on your actual experience rather than industry averages. Carriers doing this systematically often see loss ratios drop by 3 to 5 points while growing new business in the right segments [McKinsey Global Insurance Report, 2022]. Perceptive Analytics builds the data foundations that make this kind of segmentation reliable — connecting policy, claims, billing, and exposure data into a governed analytical layer that actuarial and underwriting teams can trust. Our data observability as foundational infrastructure article explains the monitoring discipline that keeps those foundations accurate over time.

Claims Analytics Deployment

This is primarily about triage — routing claims by complexity and severity at FNOL so that the right resource addresses each claim at the right time. You score claims at intake to identify fraud or high-severity cases early. The Insurance Information Institute estimates fraud costs P&C insurers approximately $30 billion annually. Good analytics can help carriers detect 80% of fraud cases compared to the typical 50% [Coalition Against Insurance Fraud, 2023]. Perceptive Analytics’ Power BI consulting and Tableau consulting capabilities build the real-time claims monitoring dashboards that make triage decisions visible to operations leadership as they happen — not after the fact.

Portfolio Restructuring

If a line of business is fundamentally broken, you may have to exit that market or restructure your reinsurance program. These are not quick fixes, but they are sometimes necessary. Expect this to be addressed during the diagnostic phase and scoped and priced separately from analytics implementation work.

Process Redesign and STP

Straight-through processing (STP) is one of the clearest efficiency metrics available. Automating reporting and building better dashboards can cut claims cycle times by 40% [RTS Labs, 2025]. But raising STP rates requires both process redesign and data quality investment — the two cannot be separated. Perceptive Analytics’ Power BI implementation services and Tableau implementation services sit at exactly this intersection — building the reporting and automation infrastructure that makes STP gains measurable and durable.

Approach Comparison: Analytical vs. Qualitative Consulting Methods

Method	Primary Lever	Typical Improvement Range	Data Dependency
Predictive loss ratio modelling	Loss ratio	3–5 pts loss ratio	High
Underwriting segmentation	Loss selection	1–3 pts combined ratio	High
Claims triage analytics	LAE / loss ratio	2–4 pts LAE ratio	Medium-High
Fraud detection analytics	Loss ratio	0.5–2 pts loss ratio	High
Process redesign and STP improvement	Expense ratio	1–3 pts expense ratio	Medium
Benchmarking only	Process / org	0–1 pt (variable)	Low

Perceptive Analytics Point of View
The table above reflects operational reality. The biggest wins come from methods that require heavy data. If a firm cannot work with your raw claims files at a granular level, their impact ceiling is low.
Check your data before you start. Are your records clean? Before scoping the engagement, assess your data readiness honestly: Is your claims data structured and accessible at FNOL level? Are your policy and loss data linked at the risk level? If not, identify what investment is needed in data infrastructure first — and build that investment into the engagement scope, not around it.

3. Measurement: How Progress and Impact on Combined Ratio Are Reported

You need a clear baseline before you start. Without one, you will never know if your “improvement” came from the consultants or just a quiet hurricane season.

Establishing the Baseline

The baseline is not your reported combined ratio. It is your combined ratio adjusted for one-time items, prior-year development noise, and catastrophe losses. Without this adjustment, it is impossible to attribute improvement to consulting initiatives versus market hardening or favorable weather. Insist on a documented baseline methodology that your own actuaries sign off on before the engagement proceeds. Perceptive Analytics builds the measurement tools before building the plan — because doing the math first is the only way to produce results that can actually be audited. Our how automated data quality monitoring improved accuracy and trust across systems case study documents what that baseline measurement discipline looks like in a production environment.

KPI Dashboards and Reporting Cadence

Track metrics that move daily — claim close time, underwriter pricing adherence, submission quality rates. These tell you whether you are winning before the quarterly financials appear. Perceptive Analytics’ Tableau development services and Power BI development services build the operational dashboards that surface these leading indicators in a format that underwriting and claims leadership actually check and use. Our frameworks and KPIs that make executive Tableau dashboards actionable article outlines the design principles that separate dashboards people trust from dashboards people ignore.

Benefit Attribution and Tracking

Keep a live register of every initiative and how much it has actually saved — not how much it was projected to save. Completing a process redesign and verifying that it reduced LAE by 1.2 points are two different things. Do not confuse finishing a task with saving money. The savings register should be reviewed at every steering committee meeting, with ownership assigned to named individuals for each line item.

4. Operating Model: Timeline, Phases, and Resource Commitment From Your Team

Your team’s time is the biggest hidden cost. Consultants bring the models, but your team brings domain knowledge, system access, regulatory context, and organizational authority. Both are required. Neither works without the other.

Typical Engagement Phases

Most combined ratio engagements follow a four-phase structure:

Diagnostic (6–10 weeks): Identify root causes and prioritize what to fix first. This phase should produce a savings register ranked by dollar impact, not a list of observations.

Design (8–12 weeks): Build the actual plans and business cases. Each initiative should have a named owner, a data dependency map, and a measurable acceptance criterion.

Pilot (8–16 weeks): Test the approach on one line of business before committing resources to a full rollout. Pilot results are the most reliable predictor of enterprise-scale outcome.

Scale (12–18 months): Roll out across the book. Perceptive Analytics’ Tableau expert, Power BI expert, and Looker consulting capabilities support the sustained analytics layer that keeps the program measurable and improving through the scale phase — after the initial implementation team has moved on.

Internal Resource Requirements

You will need a full-time project manager and a data lead. You also need a senior sponsor for each business function who can speak to operational detail and clear roadblocks. If you do not staff the project adequately, it will stall during implementation — typically at the data access or stakeholder alignment stage.

Data Access and IT Dependency

Data access is almost always the critical path in the first twelve weeks. If your data sits in legacy systems with limited extract capability, build four to six weeks of data preparation time into the engagement plan before the diagnostic begins. Perceptive Analytics’ data engineering consulting practice specializes in exactly this extraction and integration work — building the pipelines that make legacy insurance data accessible to analytics and modeling tools without requiring full system replacement. Our modern BI integration on AWS with Snowflake, Power BI, and AI case study documents what that integration architecture looks like in a production insurance environment.

5. Case Examples: What Measurable Improvement Looks Like

The following examples illustrate realistic improvement ranges across personal and commercial lines. These are representative of documented consulting outcomes, not projections.

Personal Lines — Auto Claims Transformation

A U.S. personal lines insurer was consistently ranked below peers on claim cycle times, with claims LAE running approximately 2 points above benchmark. The engagement focused on redesigning the auto claims process through customer experience strategy, back-end technology enablement, and operating model redesign. The transformation delivered a 3-day reduction in total loss cycle times and a $40 million reduction in operational expenses [Deloitte, 2023].

Commercial Lines — Underwriting Performance Turnaround

A midmarket North American P&C carrier had operated with a combined ratio above 100 for three consecutive years, with loss ratio deterioration concentrated in two commercial segments. After a comprehensive performance diagnostic, they exited underperforming segments, restructured their subrogation recovery approach, and redeployed analytics at scale across the claims function. Early savings from procurement and subrogation recovery were reinvested into end-to-end digitization of the claims operating model. The result was a 5-percentage-point improvement in combined ratio achieved in fewer than 12 months [McKinsey, 2020].

Perceptive Analytics Point of View
It wasn’t about sophisticated AI. It was about data plumbing. Most leakage happens because the right people don’t have clean data. Before you buy a new pricing tool, make sure your loss development is triangulated by cohort, your records are reconciled, and your exposure data is linked to claims at the individual risk level.
Perceptive Analytics focuses on the data infrastructure layer that makes analytical improvement possible and sustainable. Strategy without data infrastructure produces a plan. Data infrastructure without strategy produces dashboards nobody acts on. The combination produces a clear improvement program executed against a data foundation that can support ongoing measurement — and deliver combined ratio improvement that survives the next board review.

6. Economics: Typical Cost Structures for Combined Ratio Consulting

The insurance consulting market is growing as carriers invest more in data and analytics. Global spending on these services is expected to reach $10.5 billion by 2025. Within this market, how you structure the commercial model dictates the quality of what you get.

Time and Materials (T&M): Appropriate for early discovery work when you are not yet certain how deep the problems go. This requires strict governance. T&M without a hard cap is how data projects go over budget.

Fixed-Fee Phases: Best when scope is clear. The danger is data cleaning — firms often underestimate how long it takes to remediate legacy data, and when they get it wrong, the project slows or corners get cut.

Managed Services: Useful for keeping models and dashboards running after the primary engagement concludes. This is frequently under-discussed in early contract negotiations but ultimately determines whether your analytics stay accurate or degrade over time. Perceptive Analytics offers Tableau contractor and Tableau freelance developer options that provide flexible post-engagement resourcing without long-term commitment overhead — keeping dashboards and analytics environments maintained without requiring a full-time headcount.

Outcome-Based: You pay for real results. This is rare because both sides must agree on exactly how to measure savings. It works best for fraud detection where dollar-for-dollar recovery is directly attributable.

The Hidden Costs

For every dollar you spend on a new tool or model, plan to spend another dollar getting your team to actually use it. This covers training and workflow redesign. If a partner does not surface these costs in their proposal, they are setting you up for adoption failure.

Across our experience in data modernization programs, the three cost categories that most consistently surprise carriers are: data cleaning (which typically costs 3x more than estimated once the full scope of legacy inconsistencies is discovered); post-engagement support (underscoped because it is less visible during procurement); and internal staff time (your data, actuarial, and compliance teams will invest significant hours — that cost is real even when it does not appear on the invoice). Our controlling cloud data costs without slowing insight velocity guide provides benchmarks for scoping these categories realistically before contract signature.

Total Cost of Ownership

A complete TCO view should include: implementation fee, software licenses, a 30% contingency for data quality issues, and internal staff costs. Vendors who will not provide this level of transparency are a risk signal — not a negotiating style. Perceptive Analytics’ marketing analytics and chatbot consulting services extend the analytics investment into distribution and customer retention — where combined ratio improvement achieved through better underwriting and claims performance needs to be supported by retention capability that protects the profitable book you have rebuilt.

Ask vendors:

How much has your final cost differed from your initial quote across your last five engagements?
How do you handle it if my data is in worse shape than the discovery phase revealed?
What does it cost to maintain a broken data pipeline six months after you leave?

7. Risks, Downsides, and How to Mitigate Them

Improving a combined ratio comes with real risks. Acknowledging this is not pessimism — it is the realism that separates a project that works from one that generates an invoice and an exit deck.

Dependency and Capability Erosion

The largest structural risk is that your team never actually learns how to run what the consultants built. Models execute while the engagement is active, but knowledge does not transfer when they leave. The fix is ensuring your staff participates in the work daily — not as observers but as co-builders — and negotiating a formal handoff plan with defined competency milestones as a contract deliverable. Our CXO role in BI strategy and adoption article examines exactly how executive-led capability building prevents this dependency trap from forming.

Disruption and Change Fatigue

A combined ratio improvement program touches underwriting, claims, finance, and IT simultaneously. If you do not structure the work around your operational calendar — peak renewal season, hurricane response windows, IT freeze periods — your team will find workarounds that undermine the project from within. Build the engagement timeline around your business cycle, not the consulting firm’s resourcing schedule.

Overstated Improvement Claims

Some proposals promise ratio improvements that are not grounded in the client’s specific data environment. Do not accept any improvement claim that is not backed by a documented case study with math you can verify. The McKinsey 5-point case and the Deloitte $40 million savings case are useful benchmarks because they are documented — not because they are guarantees for your situation.

Perceptive Analytics Point of View
Overstating improvement is the most expensive risk in this category. If a project shows 3 points of improvement on a slide but your next annual report shows 1 point, your board will stop trusting these programs. That erosion of trust has a real cost — measured in deferred investment, delayed improvement, and leadership credibility.
At Perceptive Analytics, we build the measurement tools before we build the plan. Most firms do the opposite. Doing the math first — with a baseline that your own actuaries independently verify — is the only way to produce results that survive the next audit cycle.

8. Checklist: Questions to Ask Before You Sign an Engagement Letter

The questions below are not hypothetical — each one maps to a documented failure mode in underperforming combined ratio engagements. Work through them before you commit.

Category	Action
Scope	Can the firm show a written plan mapping tasks to specific ratio components — not just “improvement”?
Diagnostic	Does the plan include a baseline that your own actuaries will independently verify before any initiative is declared successful?
Data	Has the firm assessed your data quality — bordereaux integrity, policy-to-loss linkage, FNOL structure — before quoting a price?
Measurement	Is a live dashboard and savings register included as part of the contract deliverables, not a post-engagement add-on?
Track Record	Can they show examples from the last 24 months with specific ratio numbers, timelines, and client references you can contact?
Commercial	Is the payment structure — fixed fee, T&M, or gainshare — tied to actual delivery milestones, not calendar dates?
Capability	Is there a specific knowledge transfer plan to train your staff on the tools and models after the engagement concludes?
Governance	Is there a named senior sponsor on your side with authority to clear roadblocks across underwriting, claims, and IT?
People	Have you identified the internal leads from underwriting and claims who will work on this daily throughout the engagement?
Risk	Does the engagement plan respect your busy seasons, IT freeze windows, and regulatory filing deadlines?

Perceptive Analytics provides the data infrastructure, BI delivery, and analytics governance capabilities that make combined ratio consulting programs measurable and durable. Our full range of relevant services includes Microsoft Power BI developer and consultant services, Tableau developer and Tableau partner company capabilities, Power BI development services, and Tableau development services — all oriented toward the operational analytics layer where ratio improvement is either measured and sustained, or quietly eroded. Our answering strategic questions through high-impact dashboards case study and unified CXO dashboards in Tableau documentation both illustrate what that measurement layer looks like when it is trusted by the finance and underwriting teams using it.