The internal-production alternative should include current tooling, not an obsolete manual baseline.

Verisk’s 18 June 2026 XactGen Roofing material describes generating roofing line items and quantities inside Xactimate, with users reviewing and editing the result. This supports testing an assisted internal route; it does not demonstrate whole-file acceptance or a particular realised saving. (verisk.com)

Other evidence reinforces the need to examine the complete workflow:

• Review is a separately purchased service. Citizens’ December 2025 procurement paper identifies Crawford & Company for independent water-mitigation and mould-remediation estimate review, with fixed per-assignment compensation and no minimum commitment. It is evidence of a distinct review purchase—not a price benchmark for complete file production. (citizensfla.com)
• Review involves substantive technical work. Sedgwick’s Estimate Review Specialist posting describes checking estimates and supporting documentation against client and regulatory requirements. A vacancy is evidence of task content, not proof of an industry-wide staffing shortage. (sedgwick.wd1.myworkdayjobs.com)
• Suppliers are addressing hand-offs as well as drafting. Cotality’s August 2026 webinar description focuses on duplicate field-to-desktop entry; its May 2026 podcast discusses connected AI workflows. These are useful diligence topics, not independent productivity measurements. (cotality.com)

Our implication: require every proposal to beat a realistically improved incumbent process. Do not award savings for removing work that your existing tools could already address at lower transition cost.