Sycamore Specialty: Retirement Cliff, Tacit Knowledge, and a BPO Proposal

The two options answer different questions, which is why the debate is stuck.

What the BPO solves. Transaction volume and cost to process: intake, data entry, and issuance rising 15 percent a year with six junior assistants. It scales a non-differentiated activity and provides an accuracy service level. That is a real operational problem.

What the BPO does not solve. The retirement cliff. Losing three underwriters who hold the only authority for the two most profitable programs, and whose broker relationships carry 70 percent of the top-30 concentration, is a revenue and knowledge risk. Outsourcing policy issuance does nothing about underwriting judgment, program appetite, or broker loyalty. The CFO's framing treats a capacity problem as a cost problem.

What the academy does not solve. It does not relieve near-term processing volume, five trainees will not be productive for 12 to 18 months, and an academy with no documented curriculum and no senior underwriters free to teach will fail.

The defensible answer is a hybrid with sequencing. Outsource the transactional layer with strong controls, and simultaneously build internal capability for judgment: a documented curriculum, apprenticeship with the three senior underwriters, and codification of program appetite and pricing into written guidelines. The critical path is the retirement clock, not the processing cost; the three underwriters have 36 months and the trainees need most of it to reach independent authority.

Note the dependency. The BPO ramp is 12 months, so the transaction layer is also not a same-quarter fix. Neither option alone is sufficient, and choosing only one leaves a material risk unmanaged.

The objective is to make the programs runnable by someone else before the three underwriters leave, and to secure their cooperation while they are still there. Three workstreams, sequenced by the 36-month clock.

1. Codify the tacit knowledge. Interview each senior underwriter and extract the decision rules they actually apply: appetite and declination triggers, pricing and terms benchmarks for environmental and construction classes, referral thresholds, common broker negotiation patterns, and the claims scenarios that later became problems. Convert into written underwriting guidelines with an authority matrix. Test the codification by having a mid-level underwriter produce quotes from the guideline alone and comparing to the senior underwriter's judgment on a blind sample of prior submissions; the gap tells you what is still tacitness.

2. Transfer relationships, not just files. Over 24 months, shift joint broker calls and renewals so the senior underwriter introduces and then observes a named successor. Set an explicit target, for example that by month 18 the successor leads all top-30 broker renewals in the program. Capture broker feedback formally, because a broker who says the relationship transfers only on paper is telling you the plan is failing.

3. Retain through the transition. Retention of the three is a temporary but critical need. Consider a phased retirement option (part-time, consulting, or project-based), stay bonuses tied to documented knowledge transfer milestones rather than attendance, and equity or phantom equity tied to program results. A retention payment with no transfer obligation buys time without buying capability.

Succession of authority is the hard constraint. Program authority is a delegated underwriting decision requiring internal sign-off and, where relevant, carrier or reinsurer notification. Build the authority pipeline deliberately: name successors now, grant partial authority at 12 months, full authority at 24.

Governance. Report monthly against milestones to the CUO, with the authority transfer dates as the fixed points everything else supports.

Contract terms.

• Scope and boundaries: named processes, specified policy administration and rating systems, and explicit exclusion of any underwriting judgment, quoting, or coverage interpretation.
• Data protection: security standards (for example SOC 2 Type II), encryption, access controls, prohibition on offshore data replication outside approved locations, breach notification within a defined short window, and regulatory cooperation for any state examination.
• Accuracy service level with teeth: 99.2 percent is a starting point, but it must be defined - accuracy on what field set, measured how, by whom, with what sample size, and with what remedy. Include a fee-at-risk and escalation ladder, not just a reporting obligation.
• Subcontracting and personnel: notice and approval for subcontractors, background screening, training standards, and a requirement that the dedicated team does not rotate without notice, since turnover destroys tacit process knowledge.
• Transition and exit: a documented transition-in plan, and a transition-out plan with data return, knowledge transfer, and parallel running during any change of provider.
• Liability and indemnity appropriate to the errors and omissions exposure created by issuance errors.

Oversight program. Name an internal process owner and a governance committee with monthly review. Require a pilot on one program before full migration, with parallel running.

Metrics that matter. Accuracy and rework rate, transaction cycle time, backlog against volume, escalation resolution time, turnover of the vendor team, security incidents, and - most important - whether policy issuance errors reaching policyholders or brokers change. Track broker complaints separately, because a 99.2 percent accuracy rate on 30,000 transactions still means a high absolute error count.

Guardrail. Data security and regulatory accountability cannot be outsourced; Sycamore remains responsible to regulators and policyholders for what the vendor does.