Industry guide · Custom Software

Policy Management Software for MGAs: Build vs Buy, Real Costs, and What Actually Breaks

The short answer

If you are an MGA or program administrator running binding authority through carrier portals and Excel raters at $10 million or more in premium, building usually wins: a focused first release covering rating, policy lifecycle, and bordereaux typically runs $60,000 to $130,000 and ships in 12 to 16 weeks, with full multi-program platforms at $150,000 to $400,000 phased over 6 to 12 months.

Why policy administration makes or breaks an MGA

Walk the floor of a contractor GL program administrator writing $35 million in premium and you will find the real policy administration system: a workbook named GL_Rater_v14_FINAL_USE_THIS_ONE.xlsx, a carrier portal open in the next browser tab, and an Outlook folder called Endorsements Pending. An underwriter prices an artisan contractor risk in the rater, retypes 40 fields into the portal, the session times out at field 33, and she starts over. The policy number comes from a shared spreadsheet that two people have open at the same time. This is not a caricature. It is the standard stack for MGAs between roughly $10 million and $80 million in premium.

The leakage hides in plain sight. An underwriter who should clear 15 submissions a day clears 8 because half the day is rekeying. The operations lead loses the second week of every month to bordereaux. Surplus lines filings run on one person's memory of 22 state deadlines. And the rater itself, the artifact that decides whether your loss ratio holds, is a spreadsheet with no version control, no audit trail, and formulas nobody has tested since the actuary who built them left.

Agency management systems such as AMS360, Applied Epic, and NowCerts record what happened. They do not rate your programs, enforce your binding authority, or produce bordereaux in the formats your carriers demand. The gap between recording policies and administering them is exactly where a custom build earns its keep.

The Excel rater is your real product, and nobody controls it

A real failure mode: version 12 of a rater carries a broken VLOOKUP that underprices wind exposure in coastal counties. It runs for three weeks before the carrier's quarterly rate audit flags 61 policies priced below filed rates. Now you are writing a remediation memo to the carrier that grants your binding authority, and you cannot say with certainty which quotes used which version, because underwriters keep local copies on their desktops.

Off-the-shelf tools do not solve this. AMS360 and Epic are systems of record, not rating engines. Hosted rating vendors such as NetRate will build your rater, but every rate change enters a vendor queue with turnaround measured in weeks, you pay setup fees per program, and your actuary still develops the logic in Excel first, so you maintain two versions of the truth.

A custom build moves rating into effective-dated rate tables in a database. Every quote stores its inputs and the exact table version that priced it. Before any rate change deploys, a regression suite reprices the trailing quarter of quotes and diffs the premiums, so a broken lookup surfaces before it touches production. Underwriters get a clean quote screen with referral rules wired to your binding authority instead of an editable formula grid.

Quote, bind, and issue live in three systems that never agree

The quote exists in Excel, the bind lives in an email thread, issuance happens in the carrier portal, and the AMS record gets created two days later by an assistant working from the email. Then a mid-term endorsement arrives: an additional insured and a payroll increase. Someone computes the pro rata premium by hand, updates the portal, and forgets the spreadsheet, so the next bordereau is wrong and the carrier statement dispute takes a month to unwind.

The portal cannot fix this because it is the carrier's system, built for the carrier's ledger, and it starts at issuance. Your AMS cannot fix it because it hears about everything after the fact.

A custom platform gives each policy one lifecycle record with an explicit status machine: submission, quote, bind, issue, endorse, cancel, renew. Pro rata and short rate math is computed, not hand-keyed. Declaration pages and endorsement forms generate from the record with the correct edition dates. Where a carrier exposes an API, the system submits directly. Where the carrier only has a portal, it produces a field-by-field issuance sheet in the portal's own order, cutting a 25 minute rekeying session to under 5 and verifying the result against what was actually bound.

Bordereaux week eats four working days every month

Three carriers, three inherited Excel templates, premium and claims bordereaux for each. The ops lead assembles them by filtering AMS exports against the policy-number spreadsheet and portal downloads, and the 10th of the month is a hard deadline. One transposed cancellation and the carrier withholds settlement of the account current while everyone re-reconciles.

No agency management system exports in your carriers' formats, and portal downloads routinely miss endorsements and cancellations processed near the cycle boundary.

The custom answer is an append-only transaction ledger: written premium, endorsements, cancellations, taxes, and commission all land there as they happen. Per-carrier templates render straight from the ledger, net remittance comes out with commission and surplus lines taxes already netted, and the monthly close shrinks from four days to an afternoon of review. When a carrier changes its template, you change one mapping, not a month-end ritual.

Surplus lines compliance runs on one person's memory

An E&S book across 22 states means 22 tax rates, stamping fees, diligent effort affidavit rules, and filing calendars. A missed New Jersey filing costs penalties and interest; a pattern of misses puts the surplus lines license itself in question. Meanwhile an outdated form edition attaches to a policy because the forms folder on the shared drive holds three versions of the same exclusion, and now there is E&O exposure on top of the compliance problem.

Filing services such as InsCipher submit filings competently, but they sit downstream of issuance, fed by the same spreadsheets that created the problem.

A custom system puts compliance inside the bind path. Tax and stamping fees compute at quote time from per-state tables. Binding is blocked until diligent effort data is captured in states that require it. Filing batches export per state on their real calendars, and the forms library is keyed to edition dates so a superseded form cannot attach. Compliance stops depending on whether one specific person is on vacation.

New programs take two quarters to launch

A carrier offers appetite for a new class in six states with a 90 day window before they shop it to another MGA. Standing the program up on an MGA suite or a hosted rating vendor means months in configuration queues, and the window closes while you wait. The programs you never launched are the most expensive line item nobody books.

A custom platform treats a program as configuration: clone the nearest existing program, load the new rate tables and forms, map the states and taxes, and launch in weeks with the same audit trail from day one. Add a retail agent submission portal and growth stops meaning inbox triage, because submissions arrive structured instead of as PDFs attached to email.

What this costs and how long it takes

Across more than 2,000 delivered projects at Digital Heroes, a focused first release in this category runs $60,000 to $130,000 and ships in 12 to 16 weeks. That typically covers a rating engine for one program, the full quote-to-issue lifecycle, document generation, and bordereaux export. Complete platforms run $150,000 to $400,000 phased over 6 to 12 months, adding multi-program configuration, an agent portal, carrier connectivity, the surplus lines compliance module, and commission accounting.

What pushes price up in this specific category: the number of rating algorithms and states, the count of distinct carrier bordereaux formats, the size of the ACORD forms library, AL3 or IVANS connectivity, out-of-sequence endorsement handling, and whether a claims intake module rides along. A single-program GL rater is the cheap end; a five-program, 40-state book with three carrier feeds is the expensive end.

Build vs buy: the honest answer

Buy when you are a retail agency, where EZLynx or AMS360 plus carrier portals is genuinely enough, or when you run a single program under roughly $10 million in premium and the carrier's own portal carries the load. Enterprise policy administration suites like Guidewire PolicyCenter and Duck Creek are built and priced for carriers; for an MGA they are seven-figure, multi-year implementations solving problems you do not have.

Build when the signals stack up: three or more programs or carrier relationships, bordereaux measured in days, a rater incident you had to explain to a carrier, a program launch lost to a vendor configuration queue, or operations headcount growing faster than underwriting headcount. Our position after building these systems: past $25 million in premium with binding authority, the spreadsheet-and-portal stack is the most expensive option available. You are already paying for a policy administration system in salaries, premium leakage, and audit findings. You are just not getting one.

How to choose a developer for policy administration software

First, make them whiteboard the data model before you discuss price. A competent team distinguishes policy, term, transaction, and endorsement, uses effective-dated rate tables, and has an answer for out-of-sequence endorsements. A team that proposes one policies table with a status column will fail you in month four.

Second, test integration realism. Ask how they issue when a carrier has no API. The right answer keeps the system as the source of truth with a structured portal handoff, plus a roadmap for AL3, IVANS, or carrier APIs where they exist, and a filing service connection for surplus lines states.

Third, probe audit posture. You need an append-only ledger, an approval workflow on rate changes, and the ability to reproduce any historical premium calculation exactly. Your binding authority and your E&O carrier both depend on this.

Fourth, demand the right delivery shape: one program end to end in the first release, a parallel run against the Excel rater for a month, and acceptance defined as repricing the trailing 90 days of quotes with matching premiums. Code ownership, in writing, from the first commit. Any vendor who resists these four points has told you everything you need to know.

Research & sources

The evidence behind this guide

Independent findings on why this investment pays off. Every link goes to the primary source.

  1. Median SaaS spend reached $9,455 per employee, and organizations leave an average of 36% of their SaaS licenses unused. Source: Zylo (2026) →
  2. McKinsey argues software developer productivity can be measured by combining system-level metrics (DORA and SPACE) with its own outcome-oriented approach, which it reports deploying across nearly 20 tech, finance, and pharmaceutical companies - a claim that sparked significant debate in the engineering community. Source: McKinsey & Company (2023) →
  3. McKinsey found that currently demonstrated technologies can fully automate about 42% of finance activities and mostly automate a further 19%, indicating roughly 60% of finance work is technically automatable. Source: McKinsey & Company (2018) →
  4. Qualtrics research (Q3 2023 survey of ~28,400 consumers across 26 countries) estimated bad customer experiences put roughly $3.7 trillion in global revenue at risk annually, a 19% jump from the prior year's $3.1 trillion; 64% of customers say they will switch companies over poor service regardless of how much they like the product. Source: Qualtrics XM Institute (via Forbes) (2024) →
Rohan Malhotra · Enterprise Software Consultant

Rohan advises mid-market and enterprise teams on ERP, CRM and custom software, and has led delivery on dozens of business-software builds.

Writes for Digital Heroes, shipping business software for 2,000+ brands across 55+ countries since 2017.

FAQ

Frequently asked questions

How much does custom policy administration software cost for an MGA writing $20 to $50 million in premium?
Across Digital Heroes delivery work, a focused first release for one program typically runs $60,000 to $130,000 and ships in 12 to 16 weeks, covering rating, policy lifecycle, document generation, and bordereaux. A full multi-program platform with an agent portal and compliance module runs $150,000 to $400,000 phased over 6 to 12 months. At that premium volume the build usually costs less than a year of the operations headcount it replaces.
Should an MGA build custom policy admin software or buy something like Vertafore AIM?
Buy if you are a retail agency or a single-program MGA under roughly $10 million in premium, where an agency management system plus the carrier portal is enough. Build once you run binding authority across multiple programs or carriers, because MGA suites still leave rating in Excel and carrier reporting in manual spreadsheets. The concrete tipping points are bordereaux taking days, rater version incidents, and program launches stuck in vendor configuration queues.
How long does it take to replace our Excel rating sheets with a custom rating engine?
A production rating engine for one program typically ships inside a 12 to 16 week first release in Digital Heroes projects. The rate logic already exists in your Excel rater, so the work is converting it into effective-dated rate tables and proving parity by repricing your trailing 90 days of quotes until premiums match. Plan a parallel run of about a month before retiring the workbook.
How do we migrate existing policies from carrier portals and spreadsheets into a new system?
Migration is mostly extraction and reconciliation: policy data comes from AMS exports, carrier portal downloads, and the rater archive, then loads into the new policy records with original effective dates and transactions intact. Expect two to four weeks of data work for a book in the low tens of thousands of policies, and reconcile written premium totals against carrier statements before cutover. Historical PDFs attach to policy records so nothing lives only in old inboxes.
Do we own the source code if a development agency builds our policy administration system?
It should be written into the contract as work for hire, with the code in a repository you control from week one. That is the standard Digital Heroes arrangement: you own the code, the data, and the infrastructure accounts, and you can move maintenance to any team later. Walk away from any vendor proposing a license to their platform instead of ownership, because that recreates the lock-in you are trying to leave.
How does a custom policy admin system handle surplus lines taxes and state filings?
The system computes surplus lines tax and stamping fees at quote time from per-state rate tables, blocks binding until required diligent effort data is captured, and exports filing batches per state or feeds a filing service such as InsCipher. Deadlines become system tasks instead of one person's calendar. Every computed tax is stored on the transaction ledger so filings reconcile to the penny.
Can a custom system work with carrier portals that have no API?
Yes, and this is the realistic case since many carriers offer no API at all. The custom system stays the source of truth and generates a field-by-field issuance sheet matching the portal's screens, so rekeying drops from about 25 minutes to under 5 and gets verified against the bound quote. Where carriers do support APIs, AL3 batch, or IVANS, those connections replace the manual step program by program.
What does it cost to maintain a custom policy administration system each year?
Plan on 15 to 20 percent of the build cost per year, the typical band across Digital Heroes maintenance engagements: hosting, monitoring, rate table updates, form edition updates, and small features. For a $100,000 first release that is roughly $15,000 to $20,000 annually. Rate changes become configuration edits your own team can make rather than billable development work.
Will a custom-built policy admin system stand up to a carrier audit?
It stands up better than the spreadsheet stack, because auditability is designed in rather than reconstructed after the fact. Every quote stores its inputs and the rate table version that priced it, every transaction sits on an append-only ledger, and rate changes carry an approval record. When a carrier asks why a policy was priced a certain way, you reproduce the exact calculation instead of hunting through old workbook versions.
How many SaaS seats do we need before building custom becomes cheaper?
The crossover usually shows up between 20 and 50 seats on premium tiers. Salesforce Enterprise lists at $165 per user per month, so 40 users cost about $79,000 a year in subscriptions, which is real money against a custom system you would own outright. Run the comparison over three years: if subscription spend beats the build cost plus 15-20% annual maintenance, custom wins on price before you even count workflow fit.
We run everything on Airtable and spreadsheets. When is it time to go custom?
The switch usually makes sense when you hit one of two walls: Airtable's record caps (125,000 records per base on the Business plan) or logic the tool cannot express, like multi-step approvals with conditional pricing. There is also a simple cost signal: 25 people on Business at roughly $45 per seat per month is about $13,500 a year, forever, for a tool you are already fighting. Custom is worth it when the workflow is core to how you make money; for peripheral processes, staying on Airtable is the right call.
Will custom software work with the tools we already use, like QuickBooks and Stripe?
Yes, and this is one of custom software's genuine advantages: QuickBooks, Stripe, Shopify, and most mainstream business tools publish documented APIs built for exactly this. Expect each standard integration to add one to two weeks of build time, and be suspicious of any quote that lists five integrations without asking what data flows in which direction. The hard cases are legacy systems with no API, which is a question to raise in discovery, not in week nine.
What happens if I stop paying for maintenance after launch?
Nothing breaks on day one, which is what makes it dangerous. Within 6 to 18 months, unpatched dependencies accumulate known vulnerabilities, an integrated API like Stripe ships a breaking change, and the first fix requires a developer to relearn a stale codebase at full price. Budget 15 to 20% of the build cost per year for upkeep; it is the difference between a $500 patch and a $15,000 emergency.
Couldn't I just build my app in Bubble or another no-code tool instead of hiring an agency?
For validating an idea with real users, yes, and we tell clients that honestly. The walls come later: Bubble apps cannot be exported as code to run anywhere else, performance drops on complex data operations, and usage-based pricing climbs as you grow. A meaningful share of Digital Heroes custom builds are rebuilds of no-code MVPs that proved the business worked, which is the system operating as intended: validate cheap, then build the version that scales.
Is it cheaper to customize Salesforce than to build a custom CRM from scratch?
If you use less than a third of what Salesforce does, a custom CRM is often cheaper by year three. Salesforce Enterprise lists at $165 per user per month, so 25 seats cost about $49,500 a year before admin and consultant fees, while a focused custom CRM runs $60,000 to $100,000 once plus 15 to 20% a year in maintenance. If you genuinely need Salesforce's ecosystem, reporting, and app marketplace, customizing it beats rebuilding it; the mistake is paying enterprise prices to use it as a glorified contact list.
What are the biggest mistakes first-time software buyers make?
Choosing the lowest bid, paying more than 30-40% upfront instead of on milestones, skipping a written specification, and having no maintenance plan for after launch. The most expensive of the four in Digital Heroes rescue projects is the missing spec: without written acceptance criteria, done becomes an argument instead of a checklist, and every disagreement resolves in the vendor's favor. Fix those four and you have avoided most of the ways these projects fail.
Does the tech stack matter, and which one should I ask for?
It matters less than agencies imply, provided it is boring. A mainstream stack, something like React or Next.js on the front end, Node.js or Python behind it, and PostgreSQL for data, means thousands of developers can maintain your system if you ever change vendors. Apply one test: ask how hard it would be to hire a replacement developer for the proposed stack, and walk away from anything built on an agency's in-house framework.
Keep reading
let's build

Build something worth launching.

A plan, a team, a timeline, within 24 hours. No decks, no discovery calls. Tell us what you're building and we'll come back with a real scope and a real number.

message us directly · we reply within one business day

mission briefing

Monthly dispatch

Playbooks, real build costs, and what we're shipping. One email a month. No fluff.

visit us

New York HQ

1140 Broadway, Suite 704 · New York, NY 10001

Get directions
Online now

Hey there 👋 How can we help you today?