Policy administration is the decision that determines everything else

MGA policy administration systems sit at the centre of how a modern managing general agents operate.

They control how products are configured, how quotes are generated, how policies are bound, how bordereaux are produced and how data flows to capacity providers and regulators. Every other technology decision an MGA makes is shaped by this one.

The MGA sector is growing fast. Regulatory expectations are rising. Capacity providers are demanding better data. And half of all MGAs have barely started modernising their technology. The MGAs that choose the right platform now will compound that advantage for years. Those that delay, or choose badly, will find it increasingly difficult to compete.

This briefing draws on published data from McKinsey, BCG, Celent, AM Best, the MGAA, Conning, Capgemini, the FCA and Lloyd’s to set out the case for why platform selection matters and what the evidence says about getting it right.

A market growing at double-digit rates

The MGAA reported that its 233 members collectively wrote £13.2 billion in GWP across the UK and Republic of Ireland in 2024, covering more than 250 product lines.¹ The ABI confirmed this figure in a February 2025 joint announcement.² By August 2025, membership had grown to 249 MGAs.¹ The broader UK market contains an estimated 350+ MGAs, placing more than 10% of the UK’s £47 billion general insurance market.³,⁴

The growth trajectory is striking. From 40 members writing £1.2 billion at launch in 2011, MGAA membership has grown more than sixfold and GWP more than tenfold.¹ MGAA CEO Michael Keating stated in August 2025: “The MGA sector continues to flourish in the UK, ROI, Europe and Internationally, it remains the fastest growing P&C segment.”¹

The US and global picture is equally strong

AM Best reported that US MGA premium grew 15% year-over-year to $89.9 billion in 2024, the fourth consecutive year of double-digit increases.⁵ Conning estimated the total US market at $114.1 billion in direct premiums written.⁶ More than 700 unique MGAs met the NAIC reporting threshold, up approximately 100 from the prior year, with 19 producing $500 million or more in DPW.⁵

Globally, Insuramore reported that the MGA sector generated over $250 billion in written premiums and $29.25 billion in revenue in 2024, with approximately 3,000 MGA-related enterprises worldwide.⁷ Milliman estimated annual growth rates exceeding 20% between 2018 and 2023.⁸

Carriers are increasing their appetite for MGA capacity

Lloyd’s reported £55.5 billion in gross written premium for full-year 2024, with delegated authority accounting for approximately 45% of premium income.⁹,¹⁰ That implies £21.6 to 25.0 billion flowing through delegated arrangements including MGA coverholders. Lloyd’s maintains a global network of over 4,000 local coverholders.¹¹

A 2025 survey by Clyde & Co and the MGAA found that 57% of carriers plan to increase MGA capacity in the next two years, with 46% having already done so in the prior 12 months.¹² AM Best noted that non-exclusive MGA arrangements grew to 57% of US P&C direct premiums written in 2024, up from 33% in 2017.⁵ Lauryn Kothavale, VP for Insurance Research at Conning, noted: “Powered by talent, tech, and smart capital, MGAs continued to outpace the market in 2024.”⁶

This is the context in which an MGA chooses its technology. The sector is growing, carriers want more MGA relationships and the volume of business flowing through delegated authority is rising. The question is whether an MGA’s platform can keep pace.

Carriers are trapped by their own technology

Global insurance IT spending reached approximately $210 billion in 2023, with BCG projecting 9% annual growth to 2027.¹³ Those are large numbers, but the majority of that money goes to maintaining systems that already exist rather than building anything new.

Celent and multiple consultancies consistently report that insurers allocate between 70% and 80% of their IT budgets to maintaining legacy systems, leaving only 20 to 30% for innovation.¹⁴ McKinsey cited one insurer running more than 300 active IT systems, with over 40% slated for decommissioning and cost ratios twice the market average.¹⁵ Tech debt amounts to an estimated 20 to 40% of the value of an insurer’s entire technology estate before depreciation.¹⁶

Subhasis Mukherjee, Managing Director at Deloitte Consulting, put it bluntly: “Insurance companies have invested billions of dollars toward digital modernization, but their customers don’t see it.”¹⁷ Matthew Smith, Global Lead for Insurance Strategy and Transformation at KPMG, added: “Big technology budgets don’t necessarily lead to big cost improvements.”¹⁸

The cost of legacy is well quantified

McKinsey’s Insurance 360° benchmarking survey found that IT’s share of total operating cost at the average P&C carrier climbed from 17% to 24% over a five-year period.¹⁹ KPMG found that insurers plan to focus around 25% of their OpEx on technology.¹⁸ Much of that spend is defensive: keeping old systems running while competitors build new ones.

Capgemini’s World P&C Insurance Report 2024 found that 54% of insurers cited inadequate access to data and 51% named legacy systems as obstacles to underwriting transformation.²⁰ Oliver Wyman wrote: “An inability to escape legacy technology has been a perennial issue for most insurers. This challenge is often quoted as the biggest reason behind the industry’s slow speed to market.”²¹

None of this needs to be an MGA’s problem. But it becomes an MGA’s problem the moment it chooses a platform designed for a carrier’s world, or tries to build one from scratch and ends up creating the same kind of technical debt that carriers are spending billions trying to escape.

MGAs have an advantage that most of them are not yet using

McKinsey observed that MGAs can make the value chain more efficient because they operate without the legacy placement platforms seen elsewhere in the sector. They tend to have lean operations because they are often smaller, younger and unencumbered by the operational complexities found at insurance carriers.⁴

That structural advantage is real. An MGA choosing a policy administration system today is not migrating from a mainframe. It is not carrying 20 years of tech debt. It is making a clean choice, and if it makes the right one, it can start operating at a level that most carriers will take years to reach.

Yet a live poll at an MGAA Market Briefing in December 2025 found that 50% of MGA senior leaders are still at the earliest stages of their digital and API transformation journey.²² Half the market has the advantage and is not using it. For the other half, that is a window.

The cloud-based ecosystem is making it easier to act

Conning’s 2023 study identified the evolution of a sophisticated ecosystem of largely cloud-based service providers as one of the most important changes over the past decade, making MGAs far easier to launch and scale.²³ Only 25% of MGAs expressed concern about underinvestment in technology in 2025, down from 34% five years earlier.⁶

McKinsey estimated that cloud migration can curb IT spending by approximately 30%.²⁴ BCG found that a modern core platform can boost revenue by 25% and accelerate new-product time to market by a factor of three to four.¹³ McKinsey also found that organisations with one customer data system had IT costs 29% lower as a percentage of GWP and operating costs 12% lower than those with multiple systems.²⁵

The returns from modernisation are well documented. A 41% reduction in IT costs per policy. Over 40% more policies per FTE. Profit margin improvements of 5 to 10% in P&C.²⁶,¹⁹ These figures come from carrier modernisation studies, where the starting point was far worse. An MGA adopting a modern platform from the outset should expect to reach these efficiency levels faster and at lower cost.

Regulation is making platform capability a compliance question

The technology an MGA uses is no longer just an operational choice. Regulatory requirements now demand specific data capture, reporting and governance capabilities that only a properly configured policy administration system can deliver.

FCA Consumer Duty. Principle 12 came into force on 31 July 2023, requiring all firms in the distribution chain to deliver good outcomes for retail customers. MGAs involved in product design are classified as manufacturers under PROD 4 and must demonstrate fair value across four outcomes: products and services, price and value, consumer understanding and consumer support.²⁷

The FCA found that “where products were developed and designed by intermediaries it was more likely that firms could not demonstrate these products delivered fair value.”²⁷ Policy administration systems must capture consumer outcome data, support annual board reporting and generate the management information required for fair value assessments. An MGA running on spreadsheets or a basic quoting tool cannot produce this reporting reliably.

The FCA is looking at delegated authority models directly

The FCA’s February 2026 Regulatory Priorities announced that from Q2 2026 it will expand its oversight review to include delegated authority models and remuneration arrangements, with findings expected early 2027.²⁸ The MGAA recommended that MGAs reassess governance and oversight frameworks, particularly around claims, outsourcing and data-driven underwriting.²⁹

The FCA will also assess how firms use AI in underwriting, claims and consumer services.²⁹ MGAs deploying AI-enabled systems must demonstrate governance and outcome monitoring. This is difficult if the policy administration system does not capture granular decision data and audit trails.

Operational resilience. PS21/3 reached full compliance on 31 March 2025, requiring firms to identify important business services, set impact tolerances and conduct scenario testing.³⁰ The FCA noted that some insurance firms did not consider consumer harm from being unable to purchase, amend or renew products.³¹ A cloud-hosted platform with built-in redundancy addresses these requirements at the infrastructure level rather than requiring an MGA to engineer resilience itself.

Lloyd’s data requirements are tightening

Lloyd’s Blueprint Two is digitising the market’s core processes, with delegated authority firmly in scope. For coverholders, this means digital onboarding through DCOM, digital binder registration and structured data submission meeting prescribed standards.

Lloyd’s Coverholder Reporting Standards v5.2, released in August 2019, remain the operative standard for premium and claims bordereaux. They require regulatory, tax, premium and claims information across dozens of fields, with additional class and territory-specific data layered on top. Lloyd’s itself notes that the standard predates significant subsequent regulatory change, including the FCA’s Consumer Duty and equivalent international rules, leaving managing agents and coverholders to add their own data capture beyond the baseline.

The direction of travel is structured, API-based submission and away from the spreadsheet-based reporting that has dominated delegated authority for years. A policy administration system with native API connectivity and structured data output is now a prerequisite for efficient Lloyd’s compliance.

Bordereaux reporting is where the gap between modern and legacy is most visible

The bordereaux is how an MGA reports binding, premium and claims activity to capacity providers. It is the single most consequential data challenge that a policy administration system must solve, and the one where poor technology is most immediately exposed.

InsTech documented in 2025 that bordereaux reporting remains a major drag on the delegated authority market. Reporting typically lags by 30, 60 or even 90 days, leaving capital partners flying blind.³² The legacy approach persists: Excel spreadsheets passed around via email, formatted differently by each trading partner, reconciled manually weeks or months after the fact.³³

AM Best warned directly about this: “Insurance carriers rely on the completeness and accuracy of premium and loss data from entities that have been delegated underwriting authority. Controls over the information systems used to populate monthly reporting to carriers is critical to the completeness and accuracy of carrier financial reporting.”³⁴

The Clyde & Co/MGAA 2025 MGA Opinion report found that 77% of MGAs say the claims process needs improvement and 91% of carriers agree.¹² AM Best’s 2025 DUAE Outlook noted that data and reporting expectations are rising, adding operational complexity.³⁵ Edin Imsirovic, Director at AM Best, stated: “The current level of capacity is likely to push rates down and make performance a more important differentiator.”³⁵

When rates are falling and performance is the differentiator, the quality of an MGA’s data is what keeps capacity providers at the table. A policy administration system that produces accurate, standardised bordereaux automatically and on demand removes the single largest friction point in the MGA-carrier relationship.

What modern MGA policy administration platforms make possible

The efficiency gains available from a modern policy administration system are not theoretical. They are documented across multiple consultancy studies and real-world deployments.

McKinsey estimated that 44% of insurance work activities can be automated and that by 2030, more than 90% of pricing and underwriting for many policies will be fully automated.³⁶ Accenture found that 40% of underwriter time is currently spent on administrative tasks.³⁷ McKinsey corroborated that 30 to 43% of skilled underwriter time is wasted on administration at firms with poor system integration.¹⁵

Markel boosted underwriting productivity by 113% (measured in GWP per FTE) through a partnership with Cytora that digitised submissions and automated pre-underwriting routing, with quote turnaround times falling from an average of one day to two hours.³⁸ An EY report citing Willis Towers Watson estimated a 35% potential increase in underwriting capacity from GenAI-enabled automation.³⁸

Straight-through processing is within reach for most MGA business

McKinsey projected that up to 95% of P&C policies could undergo straight-through processing without underwriter involvement.³⁹ Capgemini found that only 8% of insurers currently qualify as “underwriting trailblazers” actively defining automated, data-driven decisions.²⁰ For an MGA, reaching this level depends entirely on the platform. Celent noted that launching a new product on a legacy system can take 18 to 24 months versus rapid configuration on a modern platform.⁴⁰

Bain & Company found that insurers taking an end-to-end AI approach to claims achieved a 35% productivity boost and cut homeowners’ claims processing times in half.⁴¹ Accenture reported that AI systems cut errors by 55% and automation can reduce claims handling time by up to 50%.⁴²,⁴³ Only 4% of insurers have scaled AI meaningfully across claims.⁴¹ An MGA with an AI-ready platform is operating in a market where almost nobody else has caught up.

The cost case is equally clear

A Bain & Company case study documented that a new digital architecture reduced annual costs by nearly 40% across claims, operations and IT. Digital natives offer three times the productivity while scoring four to five times higher in customer NPS.⁴⁴ McKinsey found that administrative costs for greenfield operations average half those of incumbents and that unit costs in P&C could go from €28 to €16 per policy through digitisation.⁴⁵

Bain estimated that generative AI could reduce loss-adjusting expenses by 20 to 25% and leakage by 30 to 50%, creating more than $100 billion in benefits across the industry.⁴⁶ These gains require a platform that can integrate AI tooling, capture structured data and automate workflows. They are not available to an MGA operating on spreadsheets and email.

Why platform selection goes wrong

The evidence on implementation failure is sobering, but the pattern is consistent: projects fail when they involve the wrong type of platform, the wrong implementation approach, or both.

Gartner found that more than 50% of life insurance policy administration projects fail to deliver the original business case.⁴⁷ BCG documented a central European group that abandoned a cross-country platforming project after eight years with a $500 million write-off, and a southern European carrier that completed its claims programme 500% over budget.¹³ KPMG found that only 25% of insurance transformation initiatives are considered highly successful.¹⁸

These failures share common characteristics. They are large-scale, multi-year, on-premise transformation programmes at carriers. They involve extensive custom development, rigid waterfall timelines and monolithic architectures. McKinsey reported that building a custom system typically takes 5 to 10 years and legacy COTS implementation requires 3 to 5 years.⁴⁸ BCG observed: “The value at stake in a core transformation is enormous”¹³, but so is the risk of getting it wrong.

The failure modes do not apply equally to every approach

McKinsey warned that common pitfalls include picking vendors without proven scalability, underestimating integration challenges and having limited visibility into platform road maps. Building projects often lock in innovation at go-live, and when overruns occur, underdocumentation and cuts to training leave organisations unable to evolve the platform.⁴⁸

Oliver Wyman added: “Many insurers have often made a mistake when digitizing their operations, they have worked back from the technology and not the customer.”²¹

These warnings are directed at carriers wrestling with decades of accumulated complexity. An MGA choosing a modern, cloud-native SaaS platform designed for its market avoids most of these failure modes by design. There is no 5 to 10 year build. There is no monolithic architecture to untangle. The deployment timeline is weeks or months, not years, because configuration replaces custom development.

The third option most evaluations skip

The build-versus-buy framing that dominates most platform conversations leaves out the option that makes most sense for a modern MGA. Configuration sits between the two. The vendor maintains, updates and invests in the platform, while the MGA shapes it to its operating model through no-code and low-code tools rather than custom development.

Building from scratch creates the same technical debt problem that carriers are spending billions to escape. Over 50% of insurtech MGAs surveyed by ReSource Pro and InsurTech NY reported designing their own policy admin systems or assembling modular technology stacks.⁴⁹ Buying a rigid traditional platform solves the maintenance problem and creates a dependency problem in its place, where every change request joins a queue behind other clients. A configurable platform avoids both, and the depth of what can now be configured without code has changed what is possible inside a vendor relationship.

What to look for in a platform partner

McKinsey’s vendor selection framework identifies six evaluation criteria: scalability and growth readiness; flexibility and system integration through API-first design and prebuilt connectors; credibility and market presence; collaboration and product influence; functionality and feature adaptability; and third-party service network and coverage.⁴⁸

At Datos Insights’ 2025 conference, Senior Principal Jackie Morales framed the point directly: “The most successful transformations happen when carriers and technology providers operate as strategic allies rather than transactional vendors.”⁵⁰

Datos Insights also found that top-performing insurers on its Data Capability Maturity Model outperform peers by 34%, yet only 34% of insurance organisations have comprehensive data governance frameworks.⁵¹ The policy administration system is where data maturity starts. If the platform cannot capture clean, structured data at the point of transaction, no amount of analytics tooling downstream will compensate.

Technology choices have a shorter shelf life than they used to

Datos Insights reported at its 2025 conference that application life cycles are compressing to approximately five years.⁵¹ This does not mean an MGA should expect to replace its platform every five years. It means the platform must evolve continuously through vendor updates, new integrations and expanding functionality. A system that is static at go-live will be outdated before the MGA has finished embedding it.

Indranil Bandyopadhyay, Principal Analyst at Forrester, predicted that insurers will “cut back on new multiyear and complex transformation programs” in favour of “more iterative peripheral development, such as building APIs and microservices.”⁵²

The direction of the market is clear. Large-scale, rigid transformation programmes are giving way to modular, API-first architectures that can be extended and adapted over time. Oliver Wyman was explicit: “CEOs must triple down on sunsetting legacy through use of rapid transition solutions, which are now available.”²¹ Capgemini agreed: “The shift away from outdated core systems in favor of cloud-native platforms continues.”⁵³

The window closes in Q2 2026

The MGA sector grew from £1.2 billion to £13.2 billion in GWP in 13 years in the UK alone.¹ The US market reached $114 billion.⁶ Carriers are increasing capacity. Lloyd’s is moving toward API-based data submission. Half of MGA senior leaders are at the earliest stages of digital transformation.²²

The FCA‘s expanded review of delegated authority models begins in Q2 2026 and reports in early 2027.²⁸ The MGAs that start their platform decision now will have answers ready when the review lands. The ones that wait will be answering questions instead.

For those that act, the advantage compounds: better data for capacity providers, automated compliance reporting, straight-through processing for routine business and a cost structure that incumbents cannot match. For those that wait, the gap widens every quarter.

The policy administration system is the one technology decision that touches every part of an MGA’s operation. It is the decision that determines everything else.

Read the full playbook in Insurance Core System Replacement: A Practical Guide for CIOs and IT Leaders.