.avif)
Why PAS automation matters in 2026
In my experience working with US P&C carriers between $500M and $5B GWP, the same Linda story repeats in every operations review. Linda is the COO (or Head of Policy Operations) responsible for 50-300 policy ops staff. Her board wants three things from her in 2026: lower expense ratio, faster product velocity, and zero agency channel friction. PAS automation is the only lever that moves all three at once.
Last quarter I sat with the COO of a Northeast specialty auto carrier ($1.4B GWP). She had pulled a single slide for the executive team. It showed her average quote-to-bind cycle time on commercial auto: 4.2 business days. The slide next to it showed her largest insurtech competitor's published cycle time: 11 minutes. She told me, "We are losing 18% of submissions to no-decision before we even rate them." That is not a technology problem in isolation. It is a PAS automation problem - manual rating spreadsheets, paper endorsement workflows, and underwriting referrals that bottleneck on three senior staff.
The numbers from Aite-Novarica back this up. In the 2025 P&C Insurance Core Modernization Report, mid-tier carriers running legacy PAS systems reported policy issuance times of 30-90 days for commercial lines and 1-7 days for personal lines. Modern PAS platforms with automation routinely drop those numbers to 5-15 minutes for personal lines and 24-72 hours for commercial - a 90% cycle time reduction. The board does not need to understand the technology. They need to see the cycle time chart.
This article is for Linda - and for Tom (Operations Director) running the PAS automation roadmap. I will not sell you "AI-powered everything." I will walk you through the four workflows where automation actually pays back inside 18 months: quote-to-bind, endorsements, renewals, and exception handling. I will tell you where automation breaks (it does, every time), what NAIC AI governance looks like in practice, and what realistic STP rates look like for a $500M-$5B GWP carrier - not for a $20B enterprise with a 40-person data science team.
What is PAS automation? Direct answer
PAS automation is the use of rules engines, workflow orchestration, no-code product configuration, and AI/ML assistance to process insurance policy transactions - quote, bind, issuance, endorsement, renewal, cancellation - without manual intervention for the majority of cases. Modern PAS automation targets 60-85% straight-through processing (STP) for personal lines and 30-60% STP for commercial lines, with human-in-the-loop review for exceptions. For mid-tier P&C carriers, PAS automation typically reduces quote-to-bind cycle time from days to minutes and rate filing deployment from 6-12 months to 24-72 hours.
That definition matters because automation is not one thing. It is a stack: a rules engine that prices and underwrites within risk appetite, a workflow engine that routes exceptions to the right team, a document engine that generates policy and endorsement paperwork, and increasingly an AI layer that pre-fills, classifies, and flags anomalies. Each layer has its own ROI curve. Skipping any layer leaves measurable money on the table.
Section 3: The four automation workflows that move the needle
In every PAS replacement and modernization project I have worked on, the same four workflows account for 80% of the operational benefit. I recommend prioritizing them in this order, because each builds on the previous one.
Quote-to-bind automation: minutes, not days
Quote-to-bind is the highest-value workflow because it directly affects agency channel productivity and submission-to-bind conversion. In a typical legacy environment, a commercial auto submission takes 30-90 minutes of producer time to enter, 4-24 hours of rater turnaround, and a 1-3 day underwriting referral cycle. With a modern PAS, an in-appetite submission gets a quote in under 60 seconds, binds in under 5 minutes, and issues policy documents within the same session.
The math: if an agency producer averages 8 submissions per day at 35 minutes each, that is 4.7 hours of data entry per day. Drop the average to 6 minutes per submission with pre-filled data from the agency management system and you free up roughly 3.9 hours per producer per day. That is real capacity - either more submissions written or fewer producers needed to maintain volume.
Endorsement automation: real-time mid-term changes
Endorsements are the operational tax of insurance. A mid-tier P&C carrier writing 200,000 policies per year typically processes 50,000-100,000 endorsements - vehicle additions, address changes, coverage adjustments, named insured updates. Legacy endorsement processing runs 3-7 days because it touches PAS, billing recalculation, document generation, and agency notification. Modern automated endorsement processing handles 80%+ of cases (vehicle add, address change, simple coverage change) in real time, generating documents, prorating premium, and updating billing in a single transaction.
Renewal automation: pre-approval batch workflows
Renewal cycles are 6-12 months out and represent 70-80% of annual premium for most P&C carriers. Automation here is less about real-time and more about batch processing - running re-rating against current loss experience, applying state-filed rate changes, flagging cases that need underwriting touch, and generating renewal documents weeks before renewal date. A well-automated renewal pipeline lets a single underwriter manage 3-5x more renewals than a manual process.
Exception handling: human-in-the-loop, not human-instead-of-loop
The 15-40% of submissions and endorsements that do not auto-process are not failures - they are the cases that need human judgment. Modern PAS automation routes these to the right person with the right context: prior history, comparable risks, underwriting guidelines, and a recommended action. The goal is not 100% STP. The goal is to route exceptions efficiently so underwriters and CSRs spend their time on judgment, not on data entry.
Straight-through processing (STP) targets for mid-tier carriers
When vendor sales decks claim "90% STP," ask which line, which submission source, and what counts as a touch. In my last 12 PAS implementations, the realistic STP rates for mid-tier P&C carriers ($500M-$5B GWP) break down as follows.
Stretch targets require deep agency data integration, third-party data enrichment (MVR, CLUE, ISO HazardHub for property, NCCI for workers comp), and an AI/ML layer that pre-classifies risk. I would not recommend any carrier target stretch numbers in year one. Get to realistic targets first. The implementation reality is that data quality - not technology - caps your STP rate. If your agency partners send incomplete submissions, no rules engine can fix that without enrichment.
I have worked with carriers who hit 78% STP on personal auto renewals in their second year post-PAS replacement, and others who plateaued at 52% because their book has a 35% non-standard auto exposure that always needs underwriting touch. There is no universal number. The right STP target for your book is the number that maximizes profit per underwriter hour, not the number on the vendor's slide.
No-code rules engine - rate filing to deployment in 24-72 hours
The single largest competitive disadvantage I see at mid-tier carriers in 2026 is the gap between state DOI rate filing approval and the rate being live in production. On legacy PAS, this gap is 6-12 months. A rate change gets approved in March, IT prioritizes it in May, gets developer time in August, tests in October, and goes live in December. By the time the new rate is in market, loss experience has moved again. The carrier is always pricing on stale data.
Modern PAS rules engines change this. A no-code rules engine lets a product manager or actuary configure rate changes, validate against test policies, and deploy to production in 24-72 hours after DOI approval. The engine handles versioning (so policies bound under the old rate continue to renew at the old rate), state-by-state variation, and time-based effective dates. In my experience, this single capability - rate-to-deployment compression - is worth 3-5 loss ratio points over a 3-year window because the carrier can react to loss experience instead of accepting it.
Configuration vs customization - the line that matters
I disagree with vendors who claim "configuration handles everything." In practice, configuration handles 70-80% of business rules. The last 20-30% - exotic commission structures, multi-line program business, niche state DOI quirks - requires custom code in some form. The honest framing: a modern PAS rules engine should let your business users handle the 80%, and your IT team should focus on the 20%. If your vendor is telling you 100% no-code, ask for three customer references at your scale running their business that way. I have not yet seen that scenario in mid-tier P&C.
AI-assisted underwriting inside PAS - governance under NAIC
AI in PAS automation is the topic that gets the most hype and produces the most disappointment. In my last 24 months, I have seen 20+ AI/ML pilots inside PAS workflows at mid-tier P&C carriers. Three made it to production at scale. The pattern of failure is consistent: the model worked in a clean test environment but failed in production because the training data did not match the live submission distribution, the model drifted within 6 months, and nobody had a monitoring framework to catch the drift.
The AI use cases that actually work in PAS automation in 2026 are narrower than vendor decks suggest. The realistic production-ready use cases I have seen:
- Document classification and extraction - taking a PDF declaration page or loss run and extracting structured data for the rating engine. Mature, low-risk, 85%+ accuracy is achievable with proper QA.
- Pre-fill from third-party data - using a license plate or address to pre-populate a submission with MVR, CLUE, or property characteristics data. Reduces producer entry time 60-80%.
- Anomaly detection on endorsements - flagging endorsements that fall outside normal patterns (e.g., a 50% premium reduction or coverage change that exceeds appetite). Catches both error and potential fraud.
- Underwriting recommendation - given a risk, suggesting an appetite/refer/decline classification with reasons. This is human-in-the-loop, not human-replaced.
The NAIC Model Bulletin on the Use of Artificial Intelligence Systems by Insurers, adopted in 2023 and now reflected in most state DOI guidance, sets the compliance bar. Carriers using AI in PAS workflows need a documented AI governance framework: model inventory, fairness testing across protected classes, documented decision logic, human-review pathways for adverse decisions, and audit trail for every AI-influenced decision. In my experience, the governance work is 40-50% of the AI implementation effort. Skip it and your state DOI examination becomes painful.
For Linda, the practical question is not "should we use AI?" It is "where in our PAS workflow does a narrow AI capability replace 20+ hours of weekly manual work?" That question has good answers. The "AI transforms underwriting" pitch does not, in my experience, in 2026.
Endorsement, renewal, and exception automation in practice
Endorsement automation - the 80/20 pattern
Endorsement automation works best when you classify endorsements upfront. About 80% of endorsements are administrative - address changes, vehicle adds, named insured corrections, lender changes. These should auto-process: validate against policy rules, recalculate premium, generate documents, update billing, notify agency. About 15% are coverage changes that need underwriting review. About 5% are mid-term cancellations or significant exposure changes that need senior underwriting. Modern PAS automation routes each into the right lane on submission, not after a manual review.
Renewal automation - 90 days out, not 30
Renewal automation should start 90 days before renewal date, not 30. The pipeline: at -90, the system re-rates against current loss experience and state-filed rate changes. At -75, exceptions (premium changes over policy thresholds, loss-driven rate increases, exposure changes) route to underwriting. At -60, approved renewals trigger document generation. At -45, agency receives the renewal package. At -30, policyholder notification. At -15, premium billing. This staged pipeline lets a small renewal team handle 50,000+ renewals per year per FTE on personal lines.
Exception handling - context, not just routing
The mistake I see most often in PAS automation is treating exception routing as a simple queue. The exception gets routed to a team inbox; the next available person picks it up. That works for volume but destroys decision quality. Modern exception handling sends context with the case: prior loss history, comparable risks the underwriter has written, applicable underwriting guidelines, and a recommended action with reasoning. The underwriter agrees, disagrees, or modifies in one screen. Cycle time drops 40-60% versus a queue-based approach.
ROI of automation per workflow
Building the business case for PAS automation requires breaking the ROI down by workflow. Generic "PAS modernization ROI" claims do not survive a CFO review. Here is the framework I use with mid-tier P&C clients.
The total business case for a mid-tier PAS automation program typically lands at 5-year NPV of $8M-$25M depending on book size, with a payback period of 24-36 months and an IRR of 18-28%. These numbers are not vendor promises. They are what I have seen across 12+ implementations. Carriers below $500M GWP struggle to justify these investments at full scope; the math gets better as book size scales.
Decerto reference cases
Three Decerto reference cases illustrate the PAS automation outcomes mid-tier carriers can expect. I have worked directly on or alongside each of these.
Generali Group Poland - 14-month full PAS migration
Generali Group Poland is the flagship Decerto PAS reference. Over 14 months, we migrated 50M+ policy records across multiple P&C lines (auto, property, commercial) onto Decerto Higson. The cutover weekend ran 32 hours; we logged zero claim payment delays and four data reconciliation issues, all resolved before Monday open. Post-cutover, the carrier hit quote-to-bind cycle time under 8 minutes on personal auto (legacy: 45+ minutes) and processed endorsements within 24 hours for 85%+ of cases. The Strangler Fig migration pattern - new policies on Higson, legacy book continuing to renew on the old system until natural attrition - kept agency channel productivity dip under 8% in the first 90 days.
Allianz - 20+ year partnership across multiple PAS cycles
Decerto has worked with Allianz across multiple PAS upgrade cycles since 2003. This is not a single-project reference; it is the demonstration that PAS modernization is iterative, not one-and-done. The most recent cycle focused on no-code rules engine automation for rate change deployment, reducing time-from-DOI-approval-to-live from 6+ months to under 72 hours. That capability alone changed how the underwriting team approached rate filings - they could now react to loss experience instead of accepting it.
Warta - multi-line P&C modernization
Warta's multi-line modernization addressed a common mid-tier reality: separate legacy systems per line of business creating inconsistent customer experience and duplicated maintenance cost. The consolidated PAS automation reduced policy administration headcount needs by approximately 18% while increasing policy volume processed by 30%. The pattern that worked: line-by-line migration over 24 months with parallel run periods, not a single big bang cutover.
Talk to Decerto about PAS automation
Every quarter you defer the PAS automation conversation, your agency channel loses ground to competitors with 5-minute quote-to-bind, your operations team absorbs more rework cost, and your underwriters spend more hours on data entry than on judgment. The compounding cost is real - and the gap is widening through 2026 as more carriers cut over.
A PAS demo with Decerto is not a slide deck. It is a working session where Marcin or another senior solution architect walks through your specific workflow constraints - your STP target, your line mix, your agency channel reality, your current rate-to-deployment timeline - and shows what Higson PAS automation actually does in production for carriers your size.
Decerto Higson PAS is not the right fit for $5B+ enterprise carriers - Guidewire PolicyCenter is built for that segment. We are built for mid-tier P&C carriers between $500M and $5B GWP who need modular, configurable PAS automation without enterprise complexity and budget. If you are above $5B GWP, we will tell you so.
The same approach we used at Generali Group Poland - 14-month migration, 50M+ records, zero claim payment delays, quote-to-bind under 8 minutes post-cutover - is the same approach we bring to mid-tier US P&C carriers. The pattern works.
Subscribe to receive the latest blog posts to your inbox every week.
