MGA Software - The Complete Buyer’s Guide

Decide what you need before you meet a vendor, then make every vendor prove it on your products rather than theirs. The eBook sets out an eight-week selection process. In week one, agree weighted criteria and pick two test cases: your hardest current product and your most demanding carrier's reporting format. In week two, screen the market by category and cut to three vendors. Weeks three and four are scripted demonstrations, where each vendor works through your test cases on your script with your data. Weeks five and six are a hands-on proof of concept, references and commercials follow in week seven and the decision comes in week eight.

The proof of concept is the most valuable fortnight. A platform that is configurable by business users in principle and by the vendor's consultants in practice shows itself the moment your underwriter tries to change a rating table, so time each task and note every point at which the vendor's staff take over. Put the chief underwriting officer, operations, finance, compliance and IT in the room, and invite your main capacity provider to say what it needs from your data in week one rather than at renewal. Our page on MGA software shows how Genasys meets the brief.

Five, and each was built for a different buyer. Enterprise carrier suites such as Guidewire, Duck Creek and Majesco run many of the world's largest insurers and earn their cost on large, complex books, though licences, implementations and change are sized for carriers. Agency and broker management systems such as Vertafore are built around clients, placements and broker accounting, and suit an MGA whose work is mainly distribution. No-code insurance platforms let business users configure products, rating, documents and workflow, run policy administration, claims and billing in one system and price by subscription. API-first headless cores give an engineering team a modern data model and leave the front end to the customer. Delegated authority and bordereaux tools help carriers and managing agents analyse bordereaux from many MGAs, alongside a policy platform rather than instead of one.

Most MGAs shortlist from the enterprise suites and the no-code platforms, with headless cores for engineering-led start-ups. The broker systems and bordereaux tools are usually complements to a core platform rather than alternatives. Genasys is a no-code platform, used by MGAs such as Arma Karma in the UK, and the eBook is clear about when a heavier suite is the better choice: for example, an MGA inside an insurance group that already licenses and develops on one.

The licence or subscription is one cost line of seven. A sound five-year comparison also counts implementation, data migration, integrations, change and customisation after go-live, your own team's time and the cost of getting your data back when the contract ends. Each line has somewhere it hides: price steps at volume thresholds, day rates outside a fixed scope, every new carrier bordereaux format after go-live and underwriters pulled off live business during build and testing.

The eBook's worked comparison models a £30 million MGA over five years. A platform with a heavy implementation and developer-led change costs £4 million in the model, a no-code subscription platform £1.65 million and doing nothing £2.65 million, most of it in people's time spent rebuilding bordereaux and rekeying data. The figures are illustrative, so build the model with your own quotes. Ask every vendor to price your business today and at three times its size, and to say what happens if a binder is lost and premium falls. Seat-based pricing penalises an MGA for giving brokers and TPAs access, while a fee tied to premium or policy count keeps the platform in proportion to the book.

For a first product on a configurable platform, about 100 days from signing to go-live is a realistic plan, and much less is possible. The eBook's 100-day plan configures the first product on one binder in days 0 to 30, migrates in-force policies and open claims and runs user acceptance testing in days 31 to 60 and goes live in days 61 to 100, with old and new bordereaux run in parallel for one cycle and the second product built by the MGA's own team. Genasys delivered Arma Karma's policy administration platform in one week, and a UK specialist broker went from specification to a live yacht product in ten days.

What slows an implementation down is size and control rather than technology. McKinsey and the University of Oxford found that IT projects with budgets above $15 million ran 45% over budget on average. Start with one product on one binder, agree the plan, milestones and acceptance criteria in the contract, and name a product owner in the business with authority over what is configured. Reconcile premium, claims and commission to the penny before the old system is switched off, and share sample bordereaux with each carrier early, so better data reaches them before their next renewal. The Arma Karma case study shows a phased launch in practice.

Ask questions that make the vendor show you, and ask references what broke. In a demonstration, ask to see a risk that breaches a binder limit and the referral and audit trail it triggers, one product written on two binders with different commission and tax, a mid-term adjustment on an instalment policy flowing through to broker, MGA and carrier balances and a bordereau produced in your carrier's format from data entered during the demo. Ask for the public API documentation before the demonstration, current ISO 27001, SOC 2 or Cyber Essentials Plus certificates and the last three months of uptime.

Then check the vendor as well as the software. Ask for audited accounts, the ownership structure and any change of control in the last three years, since an acquisition can change pricing, roadmap and support. Ask for the release history of the last two years alongside the roadmap, and agree in writing how your data and configuration come back if the vendor is sold, fails or leaves your market. Speak to two MGAs of your size that have been live for at least two years, and ask them how long implementation took against the plan, who built the product the vendor had not anticipated and whether they would choose the same platform again. The eBook's checklist turns all of this into a scorecard for every vendor meeting.