• contact@verticalserve.com
Home / Engineering / Post 20
Engineering Blog · Post #20

Win Probability Forecasting: How Manual UW Entries Drive Retention Pipeline Analytics by LOB, Underwriter, and Quarter

How InsightUW turns Sarah Chen's manual win probability entries on 8 expiring D&O policies into a live retention pipeline — with expected premium calculations, gap-to-target analysis against the 85% retention goal, and drill-down views that tell management exactly which renewals need attention this quarter.


The Problem

Retention forecasting at most specialty carriers is a quarterly exercise in spreadsheet fiction.

A D&O (Directors & Officers Liability) book manager oversees 140 expiring policies per year. Each quarter, they need to answer three questions: How much premium is likely to renew? Which accounts are at risk? Where is the gap between forecast and target?

Here is how it works today:

  • Quarterly spreadsheet survey. Management sends each underwriter a spreadsheet of their expiring accounts and asks them to fill in a "likelihood to renew" column. The underwriter enters "H" (high), "M" (medium), or "L" (low) — qualitative guesses that mean different things to different people. One UW's "medium" is another's "high."
  • No expected premium calculation. The spreadsheet shows expiring premium but not expected renewal premium. A $200K D&O policy with a 15% rate increase and a 60% win probability has a very different expected value than a $200K policy at flat rate with 95% win probability. The spreadsheet treats them identically.
  • Snapshot, not pipeline. The spreadsheet is a point-in-time snapshot. The day after submission, it is stale. The underwriter closes a deal or loses an account, and the spreadsheet does not reflect it until next quarter's survey.
  • No rollup capability. Management cannot see retention by LOB, by underwriter, by quarter, or by stage without manually pivoting and re-pivoting the data. "What is our D&O retention forecast for Q3?" requires 20 minutes of spreadsheet work.
  • No gap-to-target analysis. The carrier targets 85% retention. Is the D&O book on track? Nobody knows until the quarter ends and actual results are compared to the target. By then, it is too late to intervene.

The cost of blind retention management:

Problem Frequency Impact
Qualitative-only forecasts (H/M/L) 100% of carriers surveyed Cannot calculate expected premium or gap
Stale forecasts (updated quarterly) 85% of carriers Decisions based on 30-90 day old data
No per-renewal intervention signal 95% of carriers At-risk accounts not identified until lost
Retention target miss discovered post-quarter 80% of carriers No time to course-correct

The InsightUW Win Probability Pipeline

InsightUW replaces the quarterly spreadsheet with a live, underwriter-driven win probability pipeline. Each underwriter enters a numeric win probability (0-100%) and a proposed renewal premium for every expiring account. The system calculates expected premium, rolls up by LOB, underwriter, quarter, and stage, compares to the retention target, and surfaces the specific renewals where intervention will close the gap.

graph TD subgraph Input["UW Win Probability Entry"] A["UW opens renewal<br/>in InsightUW"] B["Enters win probability<br/>(0-100%)"] C["Enters proposed premium<br/>(renewal pricing)"] D["System calculates:<br/>expected premium =<br/>proposed × win% / 100"] end subgraph Pipeline["Pipeline Rollups"] E["By LOB<br/>(D&O, E&O, Cyber, etc.)"] F["By Underwriter<br/>(Sarah Chen, Mike Rivera, etc.)"] G["By Quarter<br/>(Q1, Q2, Q3, Q4)"] H["By Stage<br/>(pending, quoted, negotiating, bound)"] end subgraph Analysis["Gap-to-Target Analysis"] I["Target: 85% retention"] J["Actual forecast:<br/>sum of expected premiums"] K["Gap: target − forecast"] L["At-risk renewals<br/>sorted by recoverable premium"] end subgraph Action["Management Action"] M["Identify specific renewals<br/>where intervention closes gap"] N["Focus broker conversations<br/>on highest-impact accounts"] O["Adjust pricing strategy<br/>before quarter ends"] end A --> B B --> C C --> D D --> E D --> F D --> G D --> H E --> I F --> I G --> I H --> I I --> J J --> K K --> L L --> M M --> N N --> O

The Win Probability Entry API

Underwriters update win probability and proposed premium via the renewal detail screen in InsightUW, which calls the following API:

The Expected Premium Calculation

The core formula is simple and intentionally transparent:

This converts a qualitative judgment ("I think we'll probably win this") into a quantifiable dollar value that can be summed, compared, and tracked. Examples from Sarah Chen's D&O portfolio:

Insured Expiring Premium Proposed Premium Win % Expected Premium
Meridian Healthcare Systems $285,000 $310,000 75% $232,500
Cascade Financial Group $192,000 $192,000 90% $172,800
TechVault Solutions Inc. $340,000 $374,000 45% $168,300
Summit Ridge Capital $128,000 $140,000 85% $119,000
Westfield Manufacturing $215,000 $225,000 70% $157,500
Pacific Rim Logistics $167,000 $175,000 80% $140,000
Ironbridge Pharmaceuticals $420,000 $480,000 30% $144,000
Greenfield Energy Partners $95,000 $95,000 95% $90,250
Total $1,842,000 $1,991,000 $1,224,350

Sarah's 8 expiring D&O policies total $1,842,000 in expiring premium. Her expected renewal premium based on win probabilities is $1,224,350 — an implied retention rate of 66.5%, well below the 85% target.

The Pipeline Response API

Management accesses the full pipeline view through the GET endpoint, which returns summary, by-LOB, by-underwriter, by-quarter, and by-stage rollups in a single response.

Use Case: D&O — Sarah Chen's Portfolio

The Scenario

Sarah Chen is a Senior D&O Underwriter managing a portfolio of 8 expiring D&O policies totaling $1,842,000 in premium for Q2-2026. The carrier's retention target is 85%. Sarah needs to forecast which accounts will renew, at what premium, and where management needs to intervene.

The Forecast Entry Process (Timeline)

Date Action System Result
March 15 Sarah receives 8 renewal submissions auto-generated by the nightly scan. All 8 appear in her renewal queue. Status: renewal pending review for all 8
March 20 Sarah reviews Greenfield Energy Partners ($95K, clean account, no issues). Enters 95% win probability at flat renewal. Expected premium: $90,250
March 25 Sarah reviews Cascade Financial Group ($192K, 6-year client, no claims). Enters 90% at flat renewal. Expected premium: $172,800
March 28 Sarah reviews Summit Ridge Capital ($128K, good relationship, modest rate increase). Enters 85% at $140K proposed. Expected premium: $119,000
April 2 Sarah reviews Ironbridge Pharmaceuticals ($420K, FDA warning letter, 14% rate increase required). Enters 30% win probability. Expected premium: $144,000. System flags as highest-risk renewal.
April 5 Sarah reviews Meridian Healthcare ($285K, SEC disclosure, 8.8% increase). Enters 75% initially. Expected premium: $232,500
April 10 Sarah reviews TechVault Solutions ($340K, new CFO rebidding, 10% increase). Enters 45%. Expected premium: $168,300. System flags as at-risk.
April 14 Sarah reviews Pacific Rim Logistics ($167K) and Westfield Manufacturing ($215K). Enters 80% and 70% respectively. Expected premiums: $140,000 and $157,500
April 15 All 8 forecasts entered. Sarah's pipeline dashboard updates in real time. Total expected: $1,224,350. Implied retention: 66.5%. Gap to 85%: -$341,350.

What the Manager Sees

Sarah's manager, VP of D&O Lisa Huang, opens the pipeline dashboard and sees:

graph TD subgraph Dashboard["D&O Q2-2026 Pipeline Dashboard"] subgraph Summary["Portfolio Summary"] A["Expiring Premium: $8.42M<br/>Expected Premium: $6.15M<br/>Implied Retention: 73.0%<br/>Target: 85.0%<br/>GAP: -$1,008,800"] end subgraph View["By Underwriter"] B["Sarah Chen<br/>$1.84M expiring<br/>66.5% retention<br/>GAP: -$341K"] C["Mike Rivera<br/>$3.25M expiring<br/>77.0% retention<br/>GAP: -$260K"] D["David Kim<br/>$3.33M expiring<br/>72.8% retention<br/>GAP: -$408K"] end subgraph At Risk["Top At-Risk Renewals"] E["#1 Ironbridge Pharma<br/>$420K expiring, 30% win<br/>Recoverable: $294K"] F["#2 Tech Vault Solutions<br/>$340K expiring, 45% win<br/>Recoverable: $187K"] G["#3 Apex Industries (Kim)<br/>$380K expiring, 40% win<br/>Recoverable: $228K"] end subgraph Action["Manager Actions"] H["Schedule broker call<br/>for Ironbridge"] I["Authorize pricing<br/>flexibility for Tech Vault"] J["Review David Kim's<br/>3 unforecasted renewals"] end end A --> B A --> C A --> D B --> E B --> F D --> G E --> H F --> I G --> J

The Gap-to-Target Analysis

The 85% retention target means the D&O book needs to retain $7,157,000 of the $8,420,000 expiring premium. The current forecast shows $6,148,200 — a gap of $1,008,800.

Lisa drills into the gap analysis:

Analysis Dimension Finding Action
Biggest single gap contributor Ironbridge Pharmaceuticals ($420K expiring, 30% win probability) contributes $294K of potential recovery if won Schedule joint UW-broker meeting to discuss pricing alternatives (higher deductible, sublimits on SEC investigation coverage)
Biggest UW gap David Kim: $408K gap, 3 renewals without any forecast entered Require David to complete forecasts within 48 hours; review his negotiating-stage accounts
Worst stage conversion "Quoted" stage: 12 renewals at 65% average win probability — below historical 78% conversion Analyze whether rate increases are out of market; consider competitive intelligence review
Best opportunity 4 renewals in "verbal bind" at 95% average win probability = $931K virtually certain Expedite binding to lock in premium before quarter close
Month-level risk May shows lowest retention (71.9%) due to 2 large at-risk renewals Front-load May broker outreach in April

How Sarah Updates Win Probability Over Time

Win probability is not static. As negotiations progress, Sarah updates her entries:

sequenceDiagram participant SC as Sarah Chen (UW) participant API as InsightUW API participant LH as Lisa Huang (Manager) participant Broker as Broker Note over SC: April 2 — Initial forecast SC->>API: PUT win-probability: Ironbridge, 30% API-->>SC: Expected premium: $144,000 API->>LH: Dashboard updated: Gap = -$1,008,800 Note over SC,Broker: April 12 — Broker call Broker->>SC: "Client willing to discuss higher deductible<br/>if rate increase reduced to 8%" SC->>API: PUT win-probability: Ironbridge, 55%<br/>Proposed premium adjusted to $454,000 API-->>SC: Expected premium: $249,700 (+$105,700) API->>LH: Dashboard updated: Gap = -$903,100 Note over SC,Broker: April 18 — Terms agreed Broker->>SC: "Client accepts $500K deductible,<br/>8% rate increase. Verbal bind." SC->>API: PUT win-probability: Ironbridge, 92%<br/>Stage: verbal bind API-->>SC: Expected premium: $417,680 (+$167,980) API->>LH: Dashboard updated: Gap = -$735,120 Note over LH: Gap reduced by $273,680<br/>from single renewal intervention

This sequence shows the power of live forecasting. Lisa's intervention on April 2 (scheduling the broker call) led to a pricing discussion on April 12, which led to a restructured deal on April 18. The gap shrank by $273,680 from a single targeted intervention — identified by the system because it surfaced Ironbridge as the highest-recoverable-premium at-risk renewal.

The Pipeline Rollup Architecture

The pipeline response aggregates data across four dimensions. Each rollup is calculated in real time from the individual renewal forecasts:

graph LR subgraph Source["Individual Renewal Forecasts"] A["RNW-001: $310K proposed, 75% win"] B["RNW-002: $192K proposed, 90% win"] C["RNW-003: $374K proposed, 45% win"] D["... 31 more renewals"] end subgraph Calc["Expected Premium Calculation"] E["$310K × 75% = $232,500"] F["$192K × 90% = $172,800"] G["$374K × 45% = $168,300"] H["Sum all = $6,148,200"] end subgraph Rollup["Four-Dimensional Rollup"] I["by lob:<br/>D&O: $6.15M expected"] J["by underwriter:<br/>Sarah: $1.22M<br/>Mike: $2.50M<br/>David: $2.43M"] K["by quarter:<br/>Q2-2026: $6.15M"] L["by stage:<br/>Quoted: $1.91M<br/>Negotiating: $1.51M<br/>Verbal: $931K"] end A --> E B --> F C --> G D --> H E --> H F --> H G --> H H --> I H --> J H --> K H --> L

Rollup Dimensions Explained

Dimension Purpose Key Metric
by_lob Compare retention across lines of business. Is D&O lagging behind Cyber? Implied retention % vs. target
by_underwriter Identify which UWs are below target and which accounts are driving the gap Per-UW gap amount, at-risk renewal list
by_quarter Track seasonal patterns and plan broker outreach timing Monthly retention trajectory
by_stage Understand conversion funnel health. Are quotes converting to binds? Average win probability per stage, stage conversion rate

Renewals Without Forecasts

Three of David Kim's 12 renewals have no win probability entered. The pipeline treats these as excluded from the forecast but included in the target denominator — making the gap appear larger until forecasts are entered. The system sends a daily reminder to underwriters with unforecasted renewals inside 60 days of expiration.

Win Probability = 0% (Non-Renewal)

If the underwriter enters 0% win probability, the system records a forecast non renewal event. The expected premium is $0, and the full expiring premium counts against the retention gap. This is an intentional loss — the UW has decided not to pursue the renewal — and it is distinguished from an unforecasted renewal (no decision made).

Win Probability Changes After Bind

Once a renewal reaches bound status, the win probability is locked at 100% and the proposed premium is replaced by the actual bound premium. Historical win probability entries are preserved in the audit trail for forecast accuracy analysis.

Metrics: Before and After Win Probability Forecasting

Metric Before (Quarterly Spreadsheet) After (InsightUW Live Pipeline) Improvement
Forecast granularity H / M / L (qualitative) 0-100% numeric with expected premium Quantifiable
Forecast freshness Quarterly snapshot (30-90 days stale) Real-time (updated per renewal interaction) Always current
Time to generate portfolio forecast 2-3 hours (spreadsheet consolidation) 0 min (live dashboard) Eliminated
Gap-to-target visibility Post-quarter (too late to act) Real-time (actionable during quarter) Proactive
At-risk renewals identified Anecdotal ("I think we might lose Ironbridge") Systematic (sorted by recoverable premium) Data-driven
Manager intervention success rate 12% of at-risk renewals recovered 38% of at-risk renewals recovered 3x improvement
Forecast accuracy (predicted vs. actual retention) +/- 18 percentage points +/- 5 percentage points 72% more accurate
UW time spent on forecast surveys 45 min per quarter per UW 2 min per renewal (entered during normal workflow) Integrated

Key Takeaways

  1. Win probability is a number, not a letter. "75%" means something quantifiable: this $310K renewal has an expected value of $232,500 to the pipeline. "Medium" means nothing that can be summed, compared, or acted upon. The shift from qualitative to quantitative forecasting is the foundation of everything else.

  2. Expected premium is the only metric that matters for retention planning. A $420K renewal at 30% win probability contributes $144K to the pipeline. A $95K renewal at 95% contributes $90K. Without expected premium, management would focus on the $420K account by size alone — missing that the small account is actually more valuable to the forecast.

  3. Gap-to-target analysis turns a lagging indicator into a leading one. "We missed our 85% retention target" is an autopsy. "We are $1M short of our 85% target with 10 weeks left in the quarter, and here are the 3 renewals where intervention can close $700K of that gap" is a strategy.

  4. Live forecasts eliminate the quarterly survey ritual. The underwriter enters win probability during normal workflow — when they are reviewing the renewal, talking to the broker, adjusting pricing. There is no separate "forecasting exercise." The forecast is a byproduct of doing the work.

  5. Drill-down views drive specific action. Lisa does not need to ask "how is retention going?" She opens the dashboard, sees David Kim has 3 unforecasted renewals and Sarah's Ironbridge account is the single largest recovery opportunity, and takes two specific actions. The system converts ambiguity into a task list.


Ready to replace quarterly spreadsheet surveys with a live retention pipeline? InsightUW's win probability forecasting gives your managers real-time gap-to-target visibility and identifies the specific renewals where intervention moves the number.

See Win Probability Forecasting in Action →

See InsightUW run on your data

A 45-minute working session with a real broker email and your LOBs.

Request a demo