Personal lines renewal automation is a solved pattern: segment the book, sequence the outreach, let the system carry 100% coverage. Commercial lines is where owners get skeptical, and the skepticism is earned. A commercial renewal isn't one client deciding whether to re-up an auto policy; it's a 90-to-120-day cycle involving the insured's ops team, sometimes their CFO, your producer and account manager, multiple carriers, updated exposures, loss runs, and a remarketing decision with real judgment in it.
So here's the honest scoping: automation does not run a commercial renewal. It compresses the parts of the cycle that are clocks and checklists, which turn out to consume most of the calendar, so the humans can spend their time on the parts that are negotiations. Agencies that get the split right renew more accounts with less scramble; agencies that try to automate the judgment produce embarrassing emails to their largest clients. This article is the split.
Why commercial renewals leak differently
Personal lines leaks from silence: nobody touched the policy, it auto-renewed badly or lapsed. Commercial retention runs higher (the ranges commonly cited put commercial at 90-95% against 85-90% for personal lines), but each loss is enormous, and commercial accounts rarely leave over a single premium number. They leave over process: the renewal that started 30 days out instead of 120, the exposure update nobody collected so the quote came back wrong, the loss runs requested too late for competitive remarketing, the client who felt like a renewal transaction instead of a risk-management relationship.
Which means the automation target in commercial isn't outreach volume. It's the calendar: making sure every stage of a long cycle starts on time, with its inputs collected, for every account, including the mid-size ones that don't get the white-glove treatment your top ten accounts do. That's a coverage problem wearing a project-management costume, and coverage problems are what systems solve.
The 120-day cycle, split honestly
Stage by stage, who should own what:
| Days out | Stage | System's job | Human's job |
|---|---|---|---|
| 120 | Kickoff | Open the renewal, assemble the account picture (policies, endorsements, claims, last year's remarketing notes), schedule the cycle | Decide strategy: renew as-is, restructure, or remarket |
| 110-90 | Exposure updates | Send the update request, chase the checklist (payroll, fleet, locations, revenue), parse returned documents into structured fields, flag what changed | Interpret material changes; the conversation about the new warehouse |
| 90-75 | Loss runs & submissions | Request loss runs on time, assemble submission packets, track carrier acknowledgments | Pick the markets; frame the risk narrative for underwriters |
| 75-45 | Quoting | Chase carrier responses, normalize quotes into a comparison, flag coverage differences (not just premium) | Negotiate with underwriters; decide what to present |
| 45-30 | Presentation | Prepare the renewal summary from the comparison, schedule the meeting, send the agenda | Present, advise, handle the CFO's questions |
| 30-0 | Binding & delivery | Chase signatures, issue COIs, confirm carrier binding, deliver policies, log everything | The relationship close; the "here's what we did for you" conversation |
Read the system column top to bottom and notice what it is: requests, chases, parsing, assembly, scheduling, logging. Nothing on that list requires an insurance license, and everything on it is what makes commercial renewals start late today. Read the human column: strategy, interpretation, negotiation, advice. That column is why your commercial producer earns what they earn, and automation's gift to them is that the left column stops eating their calendar.
The failure mode to design against is the handoff: the system must know when a stage needs escalation (loss runs not returned in 10 days, a carrier declining, an exposure change above a threshold) and route it to a named human with context, not retry silently forever. In commercial, a silent retry is a client wondering why nobody called about the decline.
What to automate first (and what never)
For an agency starting from zero on commercial automation, the order that works:
- The kickoff clock. Every commercial account gets its renewal opened at 120 days, automatically, with the account picture assembled. This alone ends the started-late renewal, which is the root of most commercial losses. It's also pure calendar work with zero client-facing risk.
- The exposure-update chase. Templated request, structured checklist, polite persistence until returned, human flag on material changes. Clients tolerate (and frankly prefer) systematic chasing here; it reads as thoroughness.
- Loss-run requests and submission tracking. Deadlines and acknowledgments, the pure logistics layer.
- The quote comparison assembly. Normalizing carrier quotes into one view, with coverage-difference flags a human verifies before anything reaches the client.
- Never: the strategy call, the underwriter negotiation, the renewal presentation, or any communication about a declination, a non-renewal, or a claim. These aren't automation-hard; they're automation-inappropriate, and one templated email in this category costs more trust than four stages of good logistics earned.
Small-commercial books (the BOP-and-auto accounts under a few thousand in premium) are the exception that proves the split: they behave like personal lines economically, so they can run the fuller segmented sequence model with a producer on reply, which is usually where a commercial-heavy agency sees its first measurable win.
The AMS reality check
One practical warning before any of this gets built: commercial automation is only as good as the renewal data structure in your AMS, and commercial data is messier than personal lines in predictable ways. Expiration dates spread across policies that renew on different dates within one account. Exposure data living in attached PDFs instead of fields. Remarketing history recorded in free-text notes, if at all. Loss-run contacts stored in a producer's inbox rather than the carrier record.
Run the export hygiene pass with a commercial lens before committing to a timeline: the specific checks that matter here are account-level renewal-date completeness (every policy on the account, not just the lead line), whether exposure fields are structured or buried in attachments, and whether last year's remarketing outcome is recoverable anywhere. An agency whose commercial data fails those checks isn't blocked; it just sequences differently, and stage one (the kickoff clock) becomes the tool that forces the data cleanup account by account, one renewal cycle at a time. By the second cycle, the automation is running on data it cleaned itself.
The numbers to watch
Commercial renewal automation gets judged on cycle discipline before it gets judged on retention, because retention moves slowly on a book that already renews at 90%+. Track quarterly:
- On-time kickoff rate: share of renewals opened at 120 days. Should hit ~100% in month one; it's the system's whole first job.
- Exposure-update return time: median days from request to complete return. This is the stage clients control, and systematic chasing typically cuts it dramatically.
- Remarketing option rate: share of renewals presented with a genuine alternative when strategy called for one. Late loss runs are what suppress this today.
- Account-manager hours per renewal: the efficiency claim, measured. If it isn't falling, the system is adding steps instead of removing them.
- Retention and revenue retention, per the retention math, measured on a rolling year and segmented by account size, with realistic patience: commercial retention improvements show up in next year's number, not this quarter's.
The honest pitch, and its limits
If your agency runs meaningful commercial lines, here's the claim worth testing: the accounts you lose mostly leave over process, process is calendars and checklists, and calendars and checklists are precisely what automation never drops. The limit of the claim is equally clear: nothing in this article negotiates with an underwriter or sits across from a CFO, and any vendor telling you otherwise fails question five of the vendor evaluation.
If you want the 120-day split mapped onto your actual commercial book (which stages start late today, and what that's costing at renewal), book the audit; commercial cycle-discipline is part of the assessment for mixed books.
Commercial clients don't expect their agent to be a machine. They expect the machine parts of the renewal to never be the reason the human parts went badly.