Read more insights
How Cassidy Replaces the Renewal Marketing Grind for Benefits Brokers
Every brokerage runs the same Excel-and-Outlook playbook for renewals. Here's how an AI co-worker takes the manual data movement off your team's plate, without taking the broker out of the relationship.
Every benefits brokerage in America runs roughly the same renewal marketing playbook. Build the RFP in Excel. Email it to carriers in Outlook. Wait. Chase. Catch the proposals as they trickle back, PDFs and spreadsheets, every one in a different format. Re-key them into a comparison spread. Chase some more. Format the client deck. Re-quote. Bind. Send the win/loss notes. Log the activity.
It's the same flow the industry guides have documented for two decades, and the mechanics haven't materially changed, even though AI is now genuinely good at the tedious 80% of the job.
This is a close look at what changes when an AI co-worker like Cassidy runs that playbook alongside your team.
Part 1: The Manual Renewal Marketing Process, Step by Step
Before getting to what AI changes, let's be precise about what it's changing. Strip away the regional quirks and agency-specific templates, and commercial benefits renewal marketing comes down to eight repeatable steps.
- Intake. The account manager receives a renewal proposal, a group health questionnaire, dependent lists, and benefits enrollment reports from the client, each in a different format, from a different system.
- Census build. An analyst opens those files one by one, maps dependents to employees, normalizes ages and ZIPs, and rebuilds a clean census in the agency's standard Excel template.
- Current renewal spread. The same analyst pulls plan designs, deductibles, OOP maxes, coinsurance, copays, and rates out of the renewal PDF and types them into the spread, cell by cell.
- RFP package. Draft the cover letter. Look up the right carrier contacts, which vary by region, by line of coverage, and sometimes by office. Bundle the census, spread, GHQ, and supplemental docs. Draft an outgoing email to each carrier in Outlook.
- Going to market and follow-up. Send. Then chase. Confirm receipt. Answer clarification questions. Send nudges when underwriters go quiet.
- Proposal spreading. Carrier proposals arrive in wildly inconsistent formats. One carrier writes “$20 copay,” another writes “20,” another buries it inside a coinsurance percentage. The analyst re-keys each one into the master spread next to the current plan, standardizes the language by hand, and redoes the math with fragile formulas.
- Re-quote, last look, client presentation. Flag gaps against client requirements. Draft re-quote requests. Format the spread and exec summary for the client meeting. Iterate after feedback.
- Bind, win/loss, reporting. Send firm-offer requests, fire off win/loss notifications to every carrier, log activity, and run aggregate reporting, carrier-by-carrier increases, client savings, regional performance.
In real life, this isn't a 30-minute exercise. A single group's renewal cycle stretches across two to three weeks of calendar time, and a typical analyst has six or seven groups in flight at once.
And the time isn't going where you want it. It's going to retyping carrier proposals, hunting through a spreadsheet for the right Aetna rep in the Mid-Atlantic, drafting the same follow-up email for the fifth time this week, and reconciling formats that should have been standardized once and never touched again.
Part 2: Why Generic Automation and Generic AI Don't Solve This
Brokerages have tried automation before. Workflow tools that fire renewal reminders. AMS-embedded checklists. Generic chatbots bolted onto email. They mostly miss, because benefits marketing has nuances that off-the-shelf tools don't understand:
- Census schemas vary by carrier and by group. Mapping dependents to employees and normalizing tier structures is not a generic operation.
- Carrier proposals are formatted inconsistently. Plan-design language, rate basis (composite vs. age-banded), and funding type (fully insured vs. self-funded, with admin, stop loss, and claim liability components) all need translation.
- Carrier contact logic is regional and personal. Most agencies keep regional spreadsheets with hundreds of contacts that vary by state, by line of coverage, and sometimes by office.
- “Comparable plan” is an agency-specific judgment. What counts as a match depends on a rubric only the agency knows.
- The human still owns the relationship. Auto-sending emails to carriers is a non-starter.
A tool that doesn't understand these realities just moves bad information faster. The fix isn't a generic AI assistant. It's an AI co-worker that learns your process and runs it.
That's what Cassidy is.
Part 3: The Cassidy Co-Worker Model, End to End
The right way to think about AI in renewal marketing isn't “AI replaces the broker.” It's an AI teammate that fits inside the workflow your team already runs, takes the tedious data movement off their plates, and hands back the hours they should be spending as the human in the relationship.
Here's how Cassidy maps to the eight steps above.
Steps 1–2: Intake and census build, automated
The analyst drops the renewal proposal, GHQ, dependent list, and enrollment report into Cassidy. Behind the scenes, it follows the agency's own documented process for census construction. It isn't inventing a schema. It's executing the exact one the agency spent years refining, with the analyst's own template as the target output.
Mismatches between dependent lists and enrollment files get flagged. Missing data is surfaced as a question, never silently guessed. The result is a finished census in the agency's own format, ready for review in minutes instead of hours.
Replaced: opening five files by hand and rebuilding the census from scratch.
Step 3: Current renewal spread, pre-filled
The same engine parses the renewal PDF and pulls every field straight into the right cell of the agency's spread template. Anything ambiguous gets flagged. Anything missing gets surfaced.
And when carrier proposals show up later in inconsistent wording, Cassidy standardizes them against the agency's preferred terminology. Want flat-dollar copays? Percentages get converted. Use inverse percentages for coinsurance? Those get normalized too. The spread reads the same across every carrier, every time.
Replaced: typing deductibles and copays into cells, then cleaning up the language by hand.
Step 4: RFP package, drafted in your branding
Cover letters generate in the agency's own branded templates and voice. Cassidy looks up the right regional carrier contact, bundles the package, and drafts the outgoing emails directly in the analyst's Outlook drafts folder, signed off correctly as the broker of record. Nothing auto-sends. The analyst reviews and hits send.
The carrier-contact problem deserves special attention, because it's one of the silent productivity killers in this business. The fix is a live-synced contact directory: the agency's existing regional Excel files live in SharePoint, Cassidy reads from them in real time, and when a contact changes, the next email drafts itself with the new name. No re-uploads, no stale data, no hunting. And it's permission-aware, a West-region analyst asking for the right Aetna contact gets the West contact, not the Northeast one.
Replaced: digging through spreadsheets for the right rep and writing near-identical emails one at a time.
Step 5: Going to market without losing the thread
Cassidy holds state across the multi-week cycle. Three days later, when the analyst comes back to the same group, the system already knows the renewal date, the current plans, the carriers in market, who has responded, who hasn't, and what's still outstanding.
It also handles inbound the way analysts actually work. Forward a carrier proposal from Outlook to Cassidy, and it identifies the group, files the document, and updates the spread, no UI required.
And it chases on its own. Set a rule like “if a carrier hasn't responded by Thursday, send a follow-up,” and Cassidy schedules and drafts those nudges automatically.
Replaced: keeping six renewals straight in your head and remembering who still owes you a proposal.
Step 6: Proposal spreading, side by side, with discrepancy flagging
When a new proposal lands, Cassidy drops it into the spread alongside the current and renewal plans, with monthly and annual delta math already calculated. It flags discrepancies against the client's stated requirements and asks the analyst a clarifying question when something doesn't line up.
It also handles like-for-like matching using an agency-provided rubric. If a carrier returns 50 plans, Cassidy scores them against the current HSA, HMO, and PPO designs and surfaces the best matches first, without dropping the rest. Lower-scoring options stay further to the right in the spread, so nothing gets lost.
For more complex placements, down-market age-banded rates, up-market self-funded proposals with admin, stop loss, and claim-liability breakdowns, Cassidy follows the agency's own documented spreading methodology rather than guessing.
Replaced: hours of re-keying every proposal into the master spread and redoing the math by hand.
Step 7: Client-ready presentation
Cassidy drafts the client presentation in agency branding, mapping the deck, populating the value-plan deep dive, and producing a presentation-ready deliverable. The analyst spends their time on the story and the recommendation, not on PowerPoint alignment guides.
Replaced: an afternoon of formatting slides.
Step 8: Win/loss, filing, and aggregate reporting
Win/loss notifications to carriers are auto-drafted, not auto-sent. Files get pushed to the right client folder in SharePoint, Box, or Google Drive on request, cutting out the constant manual saving that eats into every cycle.
And, critically, aggregate reporting finally becomes possible. Because Cassidy stores structured data on every group, plan, and proposal it touches, it can answer “what's the average increase from Carrier X this quarter?” or “what's our per-employee savings across the Southeast?” in plain English, and build dashboards that refresh themselves. Every analyst effectively gets a data analyst sitting next to them.
Replaced: reporting that was technically possible but never actually happened, because nobody had the hours.
Part 4: The Design Decisions That Actually Make This Work
A lot of “AI for insurance” pitches sound identical on a slide. What separates a production-grade renewal marketing AI from a science experiment is a handful of deliberate design choices.
- It follows your process, not its own. Your templates, your cover letters, your rubrics, your standardization rules. The output looks like your best analyst wrote it, because Cassidy is running the playbook your best analyst already runs.
- It lives where your team already lives. Outlook for email drafts. Excel for spreads. Teams for chat. SharePoint for files. A browser extension means analysts can query Cassidy from inside their inbox without context-switching. Asking a 200-person marketing operation to learn a brand-new SaaS UI is how implementations die. Letting them keep using Outlook while the drafts simply appear in their drafts folder is how implementations succeed.
- Human in the loop, always. Emails draft; they don't auto-send. The analyst reviews and sends. The relationship stays human.
- Cross-team collaboration and coverage. When an analyst goes on PTO, a teammate can parachute in mid-renewal with full thread visibility. Permissions are group- and region-based, synced from the brokerage's existing SSO (Microsoft Entra, Okta, or similar). When someone leaves, access updates automatically. When someone joins, they inherit the right visibility on day one.
- Role-based permissions. Producers can't accidentally fire off a carrier email the analyst hasn't reviewed. Collaborators see what's shared with them but can't act on someone else's behalf. Cassidy can only ever do what the individual user is allowed to do.
- Long-term memory across renewal years. A year later, when the same client's renewal opens again, Cassidy remembers which carrier was selected and pulls last year's files automatically. If an analyst inherits an account from someone who left, they don't need to be briefed on the prior year's decision. It's already there.
Part 5: What Changes for the Brokerage
Take the manual data movement and follow-up overhead out of renewal marketing, and three things compound.
Cycle time collapses. A multi-day census-to-RFP becomes same-day. Spreading a new carrier proposal goes from hours to roughly a minute. The two-to-three-week cycle compresses, and analysts run more groups per quarter without adding headcount.
Down-market economics finally work. Small group (sub-50 lives) has always been hard to serve profitably, because the per-group margin can't absorb the manual hours. With Cassidy handling intake, spreading, and follow-up, small group turns from a loss leader into a real growth segment.
Producers and account managers go back to being advisors. You hire experienced benefits people for their judgment, their carrier relationships, and their consultative presence. Every hour spent retyping deductibles is an hour not spent on the work you actually hired them to do. In effect, every analyst gets their own analyst.
There's also a quality dimension that's easy to miss. Cassidy doesn't lose track during renewal season the way a human inevitably does when juggling six or seven groups at once. It knows where every renewal left off, what's missing, and what's next, and it can tell you instantly. Things stop falling through the cracks.
The Bottom Line
The renewal marketing process at most brokerages today is a monument to manual data movement: Excel, Outlook, PDFs, retype, repeat. It worked for decades because there was no alternative.
There's an alternative now. Cassidy follows your process, lives in your existing tools, drafts but never auto-sends, remembers every group across every renewal year, and quietly handles every paper cut that adds up to weeks of lost analyst time.
The brokerages that adopt this model first won't just run cheaper. They'll out-respond, out-spread, out-present, and out-advise the firms still doing it by hand.
The Excel-and-Outlook era had a good run. The co-worker era is here.
Build AI Agents that operate like your team
Cassidy Inc. © 2026
