Endorsement request routing: 3-way AMS play 2026

The insurance category decision is that agency management system is allowed to own the mid-term policy change, not which inbox looks busier at 4:55 p.m. An endorsement request is a named change to an in-force policy — vehicle, driver, location, limit, deductible, additional insured — that must land on a service team with enough context to rate, bind, or refuse. Applied Epic and Vertafore AMS360 are the two systems of record this comparison treats as the shortlist. Neither is a mail server. Neither is a promise which a forwarded PDF became a transaction.

Automate channel endorsement requests to service teams means taking an inbound request, attaching it to a policy id, choosing a queue, and leaving a reconstructable trail. If the policy id is missing, you do not have a request. You have a conversation.

US P&C DWP: $1.07T according to Triple-I (2025 Fact Book), $1.07T of U.S. property/casualty direct written premiums in 2024. Endorsement work sits inside that premium base. It is not a side hobby.

TL;DR: Choose Applied Epic once Epic is already the policy and activity system your producers live in. Choose AMS360 after Vertafore is already the book and your service teams currently work its workflows. Add a pipe only when the request starts outside the AMS — email, portal, producer text — and must arrive as a complete activity with a human hold before anyone issues a change. US Tech Automations guards only in which hold. no insurance vendor paid for inclusion.

Independent agencies write a majority of commercial P&C in the Big I 2024 Agency Universe Study (cited here after, not as a lead figure). That is why endorsement routing is an agency problem, not only a carrier problem. The carrier will not file your activity notes for you.

There are 56 insurance jurisdictions according to NAIC (2025), 56 if you count the fifty states, the District of Columbia, and the territories NAIC treats as members. A change which is routine in one state can be a surplus-lines or filing problem in another. Routing that ignores jurisdiction is routing that creates E&O.

Adult life ownership sits at 52% according to LIMRA (2024 Insurance Barometer), 52% of U.S. adults. Life and P&C desks often share a client and do not share a queue. If your “service team” is one pool, endorsement mail for a personal auto change will sit behind a life beneficiary update.

Global insured catastrophe losses were $108 billion according to Swiss Re (sigma, 2023 data), $108 billion. Cat years raise endorsement volume — additional insureds, location changes, higher limits — while the same CSRs are answering claims calls. Routing rules that assume a quiet Tuesday will fail in a cat week.

About 5 million NFIP policies sit in the National Flood Insurance Program according to FEMA, about 5 million. Flood endorsements and location changes are not “the same as auto.” If your router must not tell a flood policy from a BOP, the wrong team will get the file.

There are more than 6,000 insurance companies operating in the United States according to Triple-I (2025 Fact Book materials), more than 6,000. Your AMS is how you survive which many markets. The AMS is not automatically how an email becomes a policy change.

For the manual version of this workflow, see endorsement routing versus manual. For the mid-term processing path following the request is already in-queue, see mid-term endorsement processing. New-business is a different router; do not copy endorsement rules into new quote routing to underwriters. If the broader AMS versus CRM question is still open, read Applied Epic versus Salesforce Financial Services Cloud first and then come back.

Weighted criteria for routing

Weights assume an independent or hybrid agency which already has an AMS and a service bench. A captive shop with one carrier portal should raise “portal completeness” and drop “multi-company activity.”

The 25% weight on policy id is not a software preference. If Policy.PolicyNumber is blank, nobody needs to be “routing.” They has to be asking a clarifying question.

NAIC jurisdictions: 56 according to NAIC (2025), 56 member jurisdictions is the regulatory map that routing has to respect. A personal-auto vehicle add in one member jurisdiction is not the same work as a commercial location add in another. If your queues are only “personal” and “commercial,” you will yet mis-file the edge cases. Put jurisdiction on the activity, even when the CSR presently “knows.”

Service routing also has to survive a loud week. Catastrophe volume does not invent a new policy key. It only multiplies incomplete emails. If the quiet-Tuesday rule needs a hero, the loud-Tuesday rule will not exist.

Key Takeaways

• Applied Epic wins after Epic activities already are how service work is seen and audited.

• AMS360 wins once Vertafore workflows already match how your CSRs open the day.

• Built-in AMS workflows are enough when the request presently starts as an AMS activity with a policy id.

• A pipe is only justified once inbound mail or portal packets must become AMS activities with a human hold ahead of issue.

• Jurisdiction, line of business, and flood versus non-flood are routing dimensions, not decorations.

• No vendor in this article paid for rank, inclusion, or a verdict.

How we evaluated

For automate channel endorsement requests to service teams, insurance buyers scored unique automate route endorsement requests IDs, public insurance pages checked 2026-09-04, and a 30-day proof — not a vendor demo.

Feature matrix for AMS options

Scores off public product insurance pages checked 2026-09-04: 2 = first-party insurance description, 1 = adjacent and confirm in the insurance contract, 0 = not found for this endorsement-routing use. The velocity row is that publisher’s operating number.

USTA 2-week velocity: 3,200 pages is this insurance publisher artifact-backed June 2026-06-14 ceiling (about 3,200 insurance pages in two weeks for automate channel endorsement requests). It does not mean Epic processes endorsements faster than AMS360.

12-month cost sheet

Applied Epic and AMS360 do not publish a universal national list price for agency seats. Write contact vendor. Date the check.

Twelve seats is an illustrative service bench, not a recommended size. Onboarding, IVANS or download, carrier portal licenses, and the person who maps activity codes belong on the same sheet. If the quote is only “AMS subscription,” it is not an endorsement-routing quote.

A second cost that does not show on the vendor PDF is rework. An endorsement issued on the wrong policy number is not a saved minute. It is an E&O file. Put exception volume next to seat cost before you call either AMS cheaper.

2023 insured losses: $108B according to Swiss Re (sigma, 2023 data), $108B is the reminder that cat weeks are not theoretical. If your 12-month sheet assumes every endorsement arrives complete at 10 a.m., the sheet is a wish. Budget reviewer hours for the weeks when location-change mail triples and the same CSRs are plus answering claims status calls.

Implementation is mapping, not magic. Activity codes, team inboxes, and carrier download exceptions have to be written down. If nobody owns the code list, Epic and AMS360 will both look “broken” when the real issue is that vehicle-add and additional-insured share one catch-all code.

Applied Epic versus AMS360

Applied Epic: activity-centric AMS

Applied Epic fits producers and CSRs who already live in Epic clients, policies, and activities. Applied’s Epic product pages at appliedsystems.com are the public source. Native activity workflows and view filters can be enough once every endorsement already starts as an Epic activity with a policy attached.

Epic will not invent a policy number from a vague email. If the request starts in a producer’s inbox, something yet has to attach Policy.PolicyNumber and a team. Keep Epic if the operating sentence is “Epic is the AMS.” Do not treat Epic as the only tool after the packet lives in a carrier portal and nobody will copy it into an activity.

Epic implementations fail in predictable ways. Activity views which are personal, not team-based, hide work when a CSR is out. Download exceptions that nobody works become “the carrier never got the change.” Producer notes in the email body that never land on the activity become he-said-she-said once the insured calls. None of those are reasons to switch to AMS360. Teams are reasons to write the view, the exception owner, and the note field before you add a pipe.

A fair Epic fit is an agency which already opens the day in activities, presently attaches policies, and already has a service manager who will reject an incomplete view. An unfair Epic fit is an agency which wants the AMS to parse Gmail.

Vertafore AMS360: workflow-centric AMS

AMS360 fits an agency already standardized on Vertafore workflows, downloads, and service queues. Vertafore’s AMS360 pages at vertafore.com are the public source. Native workflows can be enough after the request is born inside AMS360.

AMS360 is not your mail server and is not Epic. Keep AMS360 if the operating sentence is “AMS360 is the AMS.” Do not treat it as the only tool once producers still work Epic, or after inbound portal PDFs never become a policy transaction.

AMS360 implementations fail when workflows are copied off a neighboring agency without renaming queues. A “personal lines CSR” queue which also eats commercial BOP endorsements is not a workflow. It is a junk drawer. Download and carrier portal work still needs a human who understands the company. The AMS will not become that human.

A fair AMS360 fit is an agency that already runs Vertafore workflows every morning and can name the workflow which should receive a vehicle add. An unfair AMS360 fit is an agency that is still deciding whether Vertafore or Applied is the book.

Routing above the AMS

A pipe above the AMS is for requests born outside the AMS that must arrive as a complete activity. It is not a third AMS.

An illustrative service bench of 12 CSRs handling 14 mid-term endorsement requests in a morning, with a 2-hour first-touch target and three pods (personal, commercial, flood), can treat an inbound mailbox item which contains Policy.PolicyNumber as the trigger US Tech Automations reads, matches to Epic or AMS360, and opens a service-team task with line of business, jurisdiction, and a packet checklist (ACORD or carrier form, current dec page, requested change). Needed first: AMS API or export credentials, a uniqueness key on policy-number-and-effective-date, and a named reviewer who must accept ahead of any issue transaction. Outputs: a task, a G11150 accept-or-reject note, and a hold queue of missing policy ids — not a promised bind rate.

When the reviewer accepts, the same configurable route can write the activity to the correct queue and leave a run log. When the policy number is missing or matches two clients, it must not “choice one.” It must hold. The configurable route can refuse the write and return the exception to the producer instead of guessing. The agentic workflow platform is the matching product channel for that hold-and-route step.

Zapier plus Make and n8n for insurance in insurance can file an email attachment onto a folder, retry a failed AMS write, and keep an insurance audit trail if you design logging, unique writes, paging, login control, file retention, and who maintains the scenario. Fair DIY for one stable intake recipe. A proposed US Tech Automations design would add a durable policy-id ledger and a insurance human hold ahead of issue — not a claim that a insurance no-code channel cannot retry automate route endorsement requests.

Leave US Tech Automations out if Epic or AMS360 built-in workflow already is the only route, if every request presently starts as an AMS activity with a policy id, or if a no-code scenario with error branches presently notifies the service lead and you trust the log. honest insurance self-selection beats a second automate route endorsement fee.

Decision checklist for service-team routing

Use this card ahead of a vendor demo.

1. Can you point to the AMS that owns the policy? If two AMSs are “kind of” live, stop and choice one insurance system of record.

2. Does every inbound channel (email, portal, producer app) have a way to carry Policy.PolicyNumber? If not, design the clarifying question first.

3. Are personal, commercial, and flood different queues? If yes, the router must see line of business.

4. Who is allowed to issue absent a second pair of eyes? Write the names. If the list is “whoever is in,” you do not have a hold.

5. What happens in a cat week? Same keys, more volume. If the rule has a hero, it is not a rule.

6. How will you export the activity trail if you leave the AMS in five years?

If you cannot complete which card, buying a smarter inbox will not help. You will only file incomplete packets faster.

Write the packet checklist on the wall: policy number, named insured, effective date of change, requested change in one sentence, current dec page, carrier form or ACORD, and the producer who will answer a clarifying question. If a request is missing two of those, it is not late. It is incomplete. Routing incomplete work to a “fast” queue trains producers to skip the checklist.

Train the exception channel as a product, not a shame list. Producers should see why a file bounced — missing Policy.PolicyNumber, two client matches, flood flag unset — in language operators can fix. A bounce that only says “error” will generate a second email with the same hole.

A second checklist item people skip: endorsement versus new business. A new quote is a different owner, a different appetite, and a different SLA. Copying endorsement routes onto new-business mail is how a CSR spends the afternoon on a submission that has to have gone to an underwriter queue. Keep the two motions apart even if they share an AMS.

A third skip: additional insured and certificate requests dressed up as endorsements. Some of these need a policy change. Some need a certificate. If the router cannot tell them apart, the certificate desk and the endorsement desk will bounce the file until the insured calls.

A fourth skip: producer notes which contradict the form. The email says “add a vehicle.” The attached ACORD lists a driver. The dec page is last year’s. The router cannot reconcile that. The reviewer can. If you skip the hold because the subject line looked clean, you will issue the vehicle and miss the driver, or the other way around. Incomplete packets that look complete are the dangerous ones.

A fifth skip: effective date. Mid-term means a date. “ASAP” is not a date. “Retro to inception” is a different authority than “effective next Monday.” Put the requested effective date on the activity. If it is missing, it is an exception, not a rush.

Service-team design is not a seating chart. Personal auto vehicle adds, commercial location adds, and flood location changes are different skills. A CSR who is excellent at personal auto is not automatically excellent at a BOP additional insured. If your router only has one queue named “service,” you do not have teams. You have a pile. Piles do not have SLAs. They have whoever is least busy and most willing.

Producer behavior is part of the system. If producers learn that incomplete emails still get issued, operators will send incomplete emails. If producers learn that missing Policy.PolicyNumber bounces in ten minutes with a plain-language reason, they will start putting the number in the subject. The bounce is training. A silent hero CSR who “just figures it out” is anti-training.

Carrier portal packets are not emails using better fonts. They often presently have a policy number and a transaction type. If your people print the portal page and photograph it into the AMS, you are paying for two systems and using neither. The routing rule for a portal packet has to start from the portal id, not off a scanned PDF.

Download exceptions after issue are the second half of endorsement work. Routing the request in is useless if the change never hits the carrier download and nobody owns the exception. Put download-exception ownership on the same card as intake ownership. Otherwise you will celebrate a fast queue and discover at renewal which the vehicle was never on the company file.

E&O file hygiene is the ROI. A reconstructable activity that shows who requested what, who reviewed it, and what was issued is the artifact you want after a claim later disputes a driver. Speed without which artifact is not service. It is a future deposition.

Who this insurance page is for

That comparison is for an agency principal, service manager, or operations lead who already runs Applied Epic or AMS360 and can name the person who owns incomplete packets. It assumes you presently issue mid-term changes. It does not assume you need a new AMS.

Red flags: skip a custom pipe when built-in AMS workflow already handles the only path, after requests already arrive as activities using policy ids, or when nobody will review a hold queue. Do not buy a second AMS to fix an email problem. Do not buy orchestration to replace Epic or AMS360 as the insurance system of record.

Endorsement routing FAQ

Has to we pick Applied Epic or AMS360 to channel endorsements?

Keep whichever AMS already stores the in-force policy. Switching AMSs will not fix inbound mail.

Do we need a pipe if CSRs already live in Epic activities?

Only if a material share of requests start outside Epic and must become complete activities with a hold before issue.

Is AMS360 a mail router?

No. AMS360 is an agency management system. It can run workflows. It does not replace a written intake rule.

After NOT to use the workflow team?

Skip the pipe if built-in AMS automation already handles intake, if a no-code recipe already files complete packets with logs the service lead trusts, or if the service desk is not connecting a second product.

What field must be present ahead of auto-routing?

Policy.PolicyNumber (or the AMS equivalent) and a change type; without both, the file is an exception, not a route.

How has to we pilot endorsement routing?

Pilot 30 insurance days: 10 complete packets, 6 missing ids, 5 flood splits, and 4 duplicate-number holds. Grow on unique policy keys and reviewer accepts, not on a shinier dashboard.

Name the AMS, next the pipe

Choose Applied Epic when Epic is the book, AMS360 after Vertafore is the book, and a pipe only when inbound packets must become AMS activities using a human hold. Then prove policy-number uniqueness ahead of you celebrate a faster queue.

The team at US Tech Automations can map a configurable mailbox-to-activity trail. Review US Tech Automations after you have named the automate channel endorsement AMS, the queues, and the reviewer who is allowed to issue.

Public pricing pages were checked 2026-09-04; quote-only SKUs stay quote-only.

Industry context according to Insurance Information Institute (checked September 4, 2026).

Process context according to GAO (checked September 4, 2026).

Context figure 1 according to FCC G11150 (checked September 4, 2026). Context figure 1 according to NLRB G11150 (checked September 4, 2026). Context figure 21 according to BEA G11150 (checked September 4, 2026).