How should the desk validate the commercial case?

Run three linked studies. First, interview consenting owners about actual scope, price, trigger and implementation capacity; record non-buyers as well as buyers. Second, collect comparable redacted proposals and invoices so initial design, recurring work and external costs can be separated. Third, follow delivered outcomes over a consistent period against the original baseline.

Publish sample definition, participation bias, missing data and calculation methods. Do not infer causation from a provider case. The 2026 Grassi/GVC case describes operational and bonding work, illustrating useful implementation evidence but not an independent measurement of tax-advisory ROI. Grassi — GVC Construction case study, 26 May 2026

For a first study, set a manageable recruitment target internally and disclose the achieved count rather than implying a representative sample. Compare sectors only when the cases are genuinely comparable. Until then, describe findings as case evidence and quote observations, not market averages.