The renewal date is the one thing everybody can see. The account behind it is the thing nobody can see in one place.
The expiry list comes out of the agency management or policy administration system. Premium movement sits in the rating engine. Claims history has to be pulled from a carrier portal or a loss run. Open service tickets live in the help desk, and the producer's read on the client lives in email, a CRM note, or somebody's memory. By the time a manager asks which renewals are at risk, the team is rebuilding each account by hand, and by then the useful moves, calling the client before they see the bill, remarketing the account, reviewing coverage that no longer fits, have far less runway than they needed.
This is written for agency and carrier teams that want renewals and retention to run as a steady weekly routine, not a scramble in the last two weeks before each policy expires.
Renewal work is defensive, timed, and repetitive
Renewal is a different job from new business, and the difference is what makes it easy to under-invest in. The account already exists. The clock is fixed by the expiry date, not by when a lead happens to come in. And the same book comes around every year, which makes it feel routine even when a quarter of it is quietly at risk.
So the job is not to notice the date. The system already shows the date. The job is to surface which renewals need a person before the window closes, and to give that person the account context while there is still time to act on it. Get that wrong and the failures are expensive in familiar ways: a double-digit rate increase goes out cold and the client shops the account; a policy auto-renews on terms that no longer match the exposure; a good account with a clean loss history gets no proactive contact and drifts to a competitor over price alone; a required non-renewal notice misses its deadline and the carrier is stuck on the risk for another term.
A quick test before you build anything
Before adding any tool, check whether the basic account picture is actually reachable:
- For each upcoming renewal, can you see the premium change, the claims context, any open service issues, and who owns the next action, in one view?
- Can at-risk accounts be separated from routine ones weeks before expiry, not days?
- Can the producer, account manager, service team, and a manager all see the same next step for an account?
- After the fact, can you explain why a specific account renewed, was rewritten, lapsed, or moved to a competitor?
If those answers are shaky, the problem is not the reminder settings. It is that the renewal date and the account story live apart.
Follow one policy from renewal notice to outcome
Before changing anything, it helps to trace how a single renewal actually moves today. In most agencies and carrier teams the path looks close to this:
- The system generates a renewal list by expiry date, usually a rolling monthly cohort.
- Someone pulls the proposed renewal premium and any term changes from the rating or policy system, and compares them to the expiring policy.
- Someone pulls the loss run or claims history from the carrier portal or claims system, if there is time.
- Service complaints, unresolved tickets, and the producer's sense of the relationship get attached from memory, if at all.
- The account owner, a producer, account manager, service representative, or underwriter, decides what to do: remarket, re-quote, review coverage, call the client, escalate, or let it renew.
- The outcome, retained, rewritten, lost, lapsed, or non-renewed, gets recorded inconsistently, so next year's cycle starts blind again.
Each of those steps is doable on its own. The cost is the reassembly. The renewal list does not carry the premium movement; the premium movement does not carry the claims context; the claims context does not carry the service history. Every handoff is a place where the account picture has to be rebuilt, and the accounts that get skipped in a busy week are rarely random. They tend to be the quiet mid-size accounts that were worth keeping.
Where the renewal cycle actually breaks
Renewal workflows tend to fail in the same predictable places. Naming them makes it easier to see which one is costing you the most.
The renewal list is just a list of dates
A calendar of expiry dates tells you what is due. It does not tell you which of those renewals should worry you. Worked as a pure date list, the team spends equal attention on a clean, price-stable account and one facing a large increase after a claim, which means the account most likely to leave gets no more attention than the one that was never going anywhere.
Premium movement is not attached to the account
The single biggest retention signal is how much the renewal premium moved and why. Rate increases, exposure changes, and claims surcharges all land in the rating system, not the renewal list. When the movement is not visible next to the account, the first time anyone notices a painful increase is often when the client calls, which is the worst moment to start the conversation.
Claims history sits in a separate export
Loss runs, open claims, and loss ratio signals shape both the renewal terms and the retention conversation, but they usually require a separate pull. If claims context only gets checked after a client complains about price, the team is always reacting late on exactly the accounts where a proactive conversation would have mattered.
Service problems never reach the renewal view
An account can be priced perfectly and still leave because a claim was handled poorly or a service request sat for three weeks. The service team usually knows which relationships are strained. That knowledge rarely reaches whoever is working the renewal, so a fixable problem turns into a lost account.
Big rate movers get treated like everything else
The accounts facing the largest increases are both the most likely to shop and the most valuable to save. When the workflow has no way to pull those to the front, they get the same generic renewal notice as a policy that barely moved, and the agency finds out it lost them from the carrier's commission report.
Policies auto-renew on stale terms
Auto-renewal keeps the book moving, but it also lets coverage drift out of step with the client's real exposure. A contractor whose payroll doubled, a business that bought a second location, a homeowner who finished a major renovation, all can roll over on last year's limits because nobody reviewed the exposure before the renewal ran.
Loss reasons are never captured in a usable form
When an account leaves, the reason is often known in the moment and then lost. Without a consistent place to record why, the team cannot tell whether it is losing accounts to price, service, coverage gaps, or a specific carrier's rate action, so next year's cycle repeats the same mistakes with no way to see them coming.
What a workable renewal cycle looks like
A better renewal workflow does not need to be elaborate. It needs the account story to travel with the date. In practice, that means turning the renewal book into a weekly queue where every account carries five things a person can see at a glance.
The first is timing: the expiry date, the notice window, and who owns the account. The second is what changed: the premium movement, the claims picture, and any exposure shift since last year. The third is what could break the relationship: open service issues, complaints, and payment problems. The fourth is the plan: the next action and who is doing it, whether that is a call, a re-quote, a coverage review, or an escalation. The fifth is the result: retained, rewritten, lost, lapsed, or non-renewed, with a reason recorded well enough to learn from next year.
None of that requires new software on day one. It requires the premium, the claims, the service history, and the owner's next step to sit in one view instead of five systems. Once that view exists, automation and AI have something stable to work with. Before it exists, they only speed up the reassembly.
Build the renewal account record before you automate anything
Automated reminders are not enough on their own. A reminder that a policy expires in thirty days does not tell you whether to worry about it. What the team needs first is a record that carries why the renewal matters, what could change the outcome, and who owns the next move. Keep it small enough that people will actually update it during a busy renewal week.
| Field | What it should answer | Illustrative example |
|---|---|---|
| Policy and account | Which policy, product, and client, expiring when, owned by whom? | Commercial package, ABC Contracting, expires March 14, owned by the commercial account manager. |
| Premium movement | How does the renewal premium compare to expiring, and why? | Up 22 percent: base rate increase plus a payroll-driven exposure change. |
| Claims context | What does the loss run show, and is anything open? | One property claim last year, closed; loss ratio still within appetite. |
| Coverage fit | Has the exposure changed enough to warrant a re-quote? | Payroll roughly doubled; current limits likely low for the new size. |
| Service signals | Anything straining the relationship right now? | One billing complaint two months ago, resolved slowly. |
| Relationship signals | How is the account owner reading the client and last contact? | No proactive contact in eight months; pays on time. |
| Risk status | Routine, watch, or at-risk, and on what basis? | At-risk: large increase plus a stale service issue. |
| Next action and owner | What happens next, by when, and who does it? | Account manager calls before the renewal notice mails, prepares one alternative. |
| Outcome and reason | What happened, and why, in a form you can sort later? | Retained after coverage review; reason code: proactive call plus revised limits. |
The record is not there to become a form nobody fills in. It is there to stop the same questions being answered from scratch every renewal cycle, and to make the reason an account stayed or left something you can count at the end of the year.
Separate the at-risk accounts before the window closes
Not every renewal needs a human touch. Most of the book will renew quietly and should. The point of triage is to spend the scarce human hours on the accounts where a conversation actually changes the outcome, and to do it early enough that the conversation is possible. A handful of signals do most of the sorting.
| Signal | What it usually means | A sensible next action | Who owns the call |
|---|---|---|---|
| Sharp premium increase | The client is likely to shop the account | Call before the notice mails, explain the drivers, prepare an alternative | Producer or account manager, with underwriting input |
| Recent or open claims | Possible surcharge or non-renewal, and a coverage conversation | Review the loss run, confirm carrier appetite before promising anything | Underwriter or account manager |
| Open service complaint | The relationship is at risk regardless of price | Resolve the issue first, then have the renewal conversation | Service lead |
| Exposure no longer matches coverage | The client may be under-insured; a re-quote is warranted | Review current exposures with the client before renewing | Account manager |
| High-value or concentrated account | A loss here hurts the book out of proportion | Senior owner, earlier and more personal outreach | Principal or team lead |
| No contact in twelve months | The relationship has gone quiet; competitor risk | Schedule a check-in ahead of the renewal, not at it | Producer |
Each signal is a prompt, not a decision. The system can surface that a premium jumped or a claim is open; whether to offer a concession, re-quote, or accept the loss is a person's call, made with the client's interests and the carrier's appetite both in view.
Give the book a working-backward calendar
The most common reason retention slips is simple: renewals get worked in the last two weeks, when the only realistic move left is to send the notice and hope. A calendar that counts back from the expiry date gives the risky accounts the runway they need. The windows below suit commercial lines; personal lines usually runs on a tighter clock, closer to thirty to forty-five days, and complex accounts may start earlier than ninety.
| Timing | What happens | Output |
|---|---|---|
| T-90 to T-60 | Pull the cohort, attach premium, claims, and service context, and flag the at-risk accounts | A triaged renewal queue |
| T-60 to T-45 | Owners reach out on at-risk accounts, underwriting confirms appetite, remarketing starts where warranted | Outreach and remarket plan underway |
| T-45 to T-30 | Present options, review coverage, resolve service issues, prepare any required non-renewal notices with compliance sign-off | Decisions and notices on track |
| T-30 to T-14 | Confirm terms, follow up on open items, chase unresponsive clients | Firm renewal or a documented loss reason |
| T-14 to expiry | Final confirmations, binding, and payment | Bound renewal or recorded outcome |
| T+7 | Record the outcome and reason, and note anything worth changing next cycle | An outcome log that feeds next year |
The calendar is not about adding meetings. It is about discovering the painful increases, coverage gaps, and unhappy clients while there is still time to do something, instead of after the account has already decided to leave.
Where AI helps inside the renewal workflow
AI earns its place here when it turns scattered account context into something a person can review in under a minute, and when it pushes the accounts that need a human to the front of the queue. Used that way, it takes the reassembly work off the team without taking the judgment.
The practical uses are narrow and useful:
- Summarize the loss run, service history, and producer notes into a short account brief the owner can read before a call.
- Extract renewal dates, premium changes, and exposure shifts from documents and system exports so they do not have to be transcribed by hand.
- Sort the cohort by risk signals so the queue surfaces the accounts that need attention, with the routine ones clearly separated.
- Draft the client email, the internal brief, or the renewal talking points for the owner to edit and approve.
- Flag the outliers: an unusual premium jump, an unresolved complaint, a missing loss run, or outreach that is overdue.
Every one of those is reading, sorting, drafting, or flagging. The line matters and it is worth stating plainly. The renewal offer, the coverage recommendation, the rate, the underwriting call, and the decision to non-renew or cancel stay with the accountable person. AI drafting a client email is fine; AI deciding what to offer the client is not. And none of this is regulatory or legal advice: notice timing, wording, and filing requirements belong with your compliance team and counsel, not with a model.
Where people still decide
It is worth being explicit about the decisions the workflow should never quietly absorb. Coverage adequacy, pricing and rating, underwriting appetite, retention offers and concessions, and the decision to non-renew or cancel are judgment calls with real consequences for both the client and the book. Regulatory notice timing and content sit with compliance. What the workflow does is get the right accounts in front of the right person early, with the context already assembled, so that person spends their time deciding rather than digging. The goal is to protect judgment, not replace it.
A worked example
The numbers here are invented to show the shape of the work. Your book will sort differently, but the pattern tends to hold.
Say a personal-lines agency with 4,000 policies renewing across the year, roughly 300 to 350 a month. Today, the agency management system prints the coming month's renewal list every few weeks. A service representative scans it, checks premiums where something looks off, calls a handful of clients whose increases seem large, and lets the rest renew. Claims history only gets pulled when a client complains about a price. Lost accounts are noticed later, when the carrier commission report shows them gone.
Now picture the same month worked as a queue. The cohort of about 320 renewals lands with premium movement attached from the rating system, a loss run summary from the carrier portal, and open service tickets from the help desk. AI drafts a one-paragraph brief for each account and sorts them into roughly 250 routine, 50 to watch, and 20 at-risk. The team spends its week on the 70 that matter. The 20 with double-digit increases get a call before the client sees the bill. The 50 on the watch list get a quick check. Two accounts turn up where the exposure has clearly changed, a client who bought a second home and a small contractor whose payroll jumped, and both get flagged for a coverage review before they roll over on stale limits. Every outcome is recorded with a reason.
The payoff shows up the following year. Because the reasons were captured, the agency can see that most of the accounts it lost followed an increase over a certain size where no one called first. That is a fixable pattern, and it only becomes visible once the losses stop being invisible.
| Renewal | Risk state | Next action | Owner |
|---|---|---|---|
| Commercial package, growing contractor | At-risk: large increase plus payroll change | Call before notice, prepare a revised-limits option | Account manager |
| Fleet policy, mid-size haulier | Watch: broker awaiting revised terms | Send the option summary and confirm appetite | Producer |
| Professional indemnity, advisory firm | Internal review: recent claim | Confirm underwriting appetite before quoting | Underwriting lead |
| Homeowners, renovated property | Coverage gap: exposure changed | Review limits with the client before renewal runs | Service representative |
What to measure
A few things are worth watching, and none of them require a heavy dashboard. Track how many renewals have an owner, a risk status, and a next action before the final two weeks, because that is the difference between a proactive book and a reactive one. Track how many at-risk accounts were contacted with runway rather than after the client had already started shopping. Watch the retention rate broken out by risk band and by reason, so a drop points somewhere specific. And keep an eye on the manual hours spent assembling renewal context each month, since that is the cost the workflow is meant to take out. Over time the reason codes matter most of all. If you cannot say why last year's losses happened, you cannot change this year's.
Common traps
Most teams that set out to fix renewals hit the same avoidable mistakes. These are different from the structural breaks above; they are the errors made while trying to improve things.
Starting with a reminder tool
Reminders tell you a date is coming. They do not tell you which renewals are risky or why, so a calendar of alerts still leaves the team reacting late on the accounts that were always going to be hard.
Automating outreach before owners agree on what at-risk means
If the team has not agreed on the risk bands and who owns each account, automated client emails will land on the wrong accounts and in the wrong tone. That erodes trust faster than a slower manual process, and it is hard to undo.
Overbuilding the record so nobody keeps it current
A record that takes ten minutes to update will not get updated during renewal week. Capture what changes the decision and leave the rest; a small record that stays current beats a thorough one that goes stale by the second cohort.
Treating premium as the only retention signal
Price matters, but plenty of accounts leave over a mishandled claim or a slow service response while the premium barely moved. If service and claims stay out of the renewal view because they live in other systems, the workflow will keep missing the real reason accounts go.
Never writing down why accounts were lost
A lost account with no recorded reason teaches nothing. Without consistent loss reasons, the team cannot separate a carrier's rate action from a service failure from a genuine coverage gap, and it will keep guessing.
The first month
The first build should make one renewal cohort easier to work, not fix the whole book at once. Start narrow enough to finish.
- Pick one cohort: a single product line, one team, or one month's renewals, and map the path from the renewal list to outcome capture.
- Define the record and the risk bands with the people who actually own the accounts, so the labels mean the same thing to everyone.
- Connect the few sources that matter most: the policy or rating system for premium movement, the carrier portal for claims, the help desk for service issues.
- Stand up a weekly queue that shows routine, watch, at-risk, waiting-on-broker, and internal-review accounts, each with an owner and a next action.
- Add AI account briefs and draft client updates where the owner stays in control of both the decision and the wording.
By the end of the month the weekly renewal review should run from the queue, not from a printed list and a set of scattered notes.
How Ubisar would implement this workflow
In week 1, Ubisar would choose one renewal cohort and follow a handful of policies from renewal notice to retained, rewritten, pending, lost, or non-renewed. The first output would be a renewal queue that shows, for each account, the owner, the risk band, the premium movement, a claims summary, any open service issues, the last contact, the next action, and the expiry date.
In weeks 2 and 3, we would connect the minimum policy, rating, claims, service, and CRM data needed to keep that queue current. AI would summarize account context and draft client and internal updates, while the account owner decides the action and records the final reason. The coverage, pricing, underwriting, and retention-offer decisions stay with the people accountable for them.
By week 4, the weekly renewal review should run from the queue instead of a calendar and scattered notes. Keep going if at-risk accounts are surfacing with real runway and lost or retained reasons are getting easier to learn from; narrow it if risk bands or owners are still inconsistent. That is the shape of the AI, Data & Tech Implementation service: pick one workflow that is slowing the business, fix the data and tools around it, and ship something the team actually uses. If renewals and retention are where your month gets eaten, get in touch and we can start with a single cohort. Useful next reads: the insurance sector page, the AI readiness assessment, the relationship and client service workflow, and the product and customer analytics workflow.