In Part 2 of a series of articles on innovation in insurance, “The Hardest Part of Innovation in Insurance Isn’t Technology; It’s Culture,” Haden Kirkpatrick offers some remedies to fix incentives, governance and rituals within insurance companies that drag down innovation efforts.

Here, he delivers two bonus tear sheets to help insurance leaders put the insights into action: a long-term playbook and a 90-day plan to start a cultural turn.

***

The Playbook

Play 1: Establish a Three-Tier Risk Model for Innovation

Explore (0–3 months)

• Spend cap: very small (<$50K per innovation)
• Goal: evidence, not revenue
• Controls: templated DPIAs (Data Protection Impact Assessments), sandbox marketing claims, synthetic data only
• Exit criteria: customer validation, feasibility proof, clear adoption path

Build (3–12 months)

• Spend cap: modest ($50K-$250K per innovation)
• Goal: repeatable unit economics signals
• Controls: limited real data with monitoring, model risk memo, second-line checklists
• Exit criteria: target CAC/LTV bands, operational playbook draft, committed sponsor

Scale (12+ months)

• Spend cap: grows with traction, based on proven unit economics; bottom-line profitability TBD based on validated models.
• Goal: integration, enablement, risk-adjusted returns
• Controls: full model validation, conduct testing, change management, audit trail
• Exit criteria: product fully onboarded to BAU; product owner embedded in line organization; organizationally aligned KPI dashboards with product-level reporting agreed

Why it works: It replaces binary “approved/denied” with a graduated path where the right controls apply to the right uncertainty.

Play 2: Convert the Annual Plan Into a Quarterly Investment Council

Council chaired by a business P&L owner, not IT, finance, actuarial or other more cost-constrained teams.

• Portfolio view shows bet size, stage, runway remaining, evidence score (stoplight) and next decision date.
• 20–30 percent of the portfolio expected to be killed annually; reward timely kills and celebrate smart failures.

Why it works: Money follows learning, not calendar. It also normalizes kill decisions.

Play 3: Make Risk and Compliance Co-Owners With SLAs

Assign a named risk partner to each product pod; outline clear expectations for service levels, cultural alignment and performance benchmarks.

• Publish a response-time charter (e.g., 80 percent of standard reviews completed within five business days).
• Introduce “approval by exception”: if the SLA window lapses, the experiment proceeds within the pre-agreed guardrails.

Why it works: Shifts risk from gatekeeper to enabler, without compromising oversight.

Play 4: Institutionalize Pre-Mortems and Red Teaming

Before building, run a one-hour pre-mortem: “It’s 12 months later and this failed. Why?”

• Use rotating red teams (actuarial, claims, compliance, distribution) to attack the approach.
• Align rotating red teams based on concepts tested (e.g., claims for claims innovations; distribution for distribution innovations, etc.); however, set expectations that their role is input, not veto.
• Capture mitigations as explicit backlog items.

Why it works: You protect downside with structured skepticism while preserving upside speed.

Play 5: Productize Adoption

Define a standard Go-to-Field Pack: value prop scripts, broker enablement materials, pricing guardrails, claims handling playbooks, operational KPIs and training modules.

• Require a “first 10 customers” plan (named brokers or partners) before build.

Why it works: You de-risk the post-pilot chasm where most good ideas die.

***

A 90-Day Plan to Start the Cultural Turn

Days 0–30: Diagnose and Set the Narrative

• Artifact review. Inventory the innovation portfolio, governance documents and metrics. Classify each initiative by stage and health.
• Voice of the middle. Interview 15–20 mid-level managers in underwriting, claims, risk, distribution, finance. Ask: What stops you from saying yes?
• Publish the “why now.” A two-page memo from the CEO/ExCo ties changes to macro forces (climate, inflation, cyber risk, distribution disruption, AI) and states the kill-rate ambition.

Days 31–60: Rewire the Operating System

• Launch the three-tier model. Codify guardrails, checklists and SLAs with risk/compliance.
• Stand up the investment council. Book the monthly cadence; force rank the portfolio; reallocate 10–20 percent of budget to the highest-evidence bets.
• Name the DRIs (directly responsible individuals). For each product, appoint a single accountable owner with a multi-year

Days 61–90: Prove the New Behavior Works

• Run two fast Explore cycles. Ship evidence, not code. Publicly celebrate one timely kill.
• Graduate one Build to Scale. Show the mechanics of adoption: enablement, training, ops readiness.
• Publish the first learning dashboard. Make it visible—wins and misses.

Days 90+: Ongoing Executive Support, Engagement and Empowerment

• Set the two-track model and clarify roles. Share broadly between “innovation” product and actuarial/underwriting product but allow no vetoes; only cross pollenate when relevant and innovations are proven.
• Celebrate innovation successes, failures and fast kills at relevant values by relevant business and P&L owners; make it a core part of corporate communications.
• Revisit regularly the “Big Why” and connect innovation efforts tactically to the overall strategic narrative.

Frequently Asked Pushbacks (and How to Answer)

• “Regulators won’t allow this.” Invite them early. Most supervisors support controlled sandboxes with clear consumer protections and monitoring. Show your tiered guardrails.
• “We can’t afford failed experiments.” You already fund them. They’re slow political projects that die after two years. Make failure fast and cheap, and harvest the learning.
• “Our people are too busy.” Busy doing what the current system rewards. Change incentives, free capacity by stopping low-yield work, and rotate your best operators into scale squads.
• “Brokers/Agents won’t buy in or push it.” Brokers and agents are coin-operated: They push what creates value and is easy to explain. Co-design the proposition with them, give them repeatable playbooks, and share upside (or reduce friction) for early adoption.