The Nature for Climate Fund commissioned five years of tree planting and peatland restoration without building outcome measurement into its design. The NAO cannot assess value for money. Defra says it expects very high value for money. Those two positions are not compatible.
Airservices signed a cost-incentive contract at twice the supplier's own fixed-price bid, then expressed surprise seven years later that the incentive model hadn't contained costs.
DfT has a well-designed lessons-learned framework. The Government Internal Audit Agency confirmed in January 2026 that programme teams don't apply it. The NAO's March 2026 prescription for Northern Powerhouse Rail is to monitor whether the lessons are being applied.
Adoption statistics measure the newest tool, not the deepest capability. The functions being graded 'low uptake' have been running AI since before the people writing the reports knew what it was.
The ANAO audit of Australia's $7 billion infantry vehicle procurement found that 'mature, proven technology' appeared in the RFT without definition, threshold, or evaluation criteria. The selected platform is developmental. The lesson is not to write better requirements. It is to stop treating a stated requirement as a substitute for a measurable one.
The OAG found Oranga Tamariki applied procurement cost-reduction logic to $500 million in children's services without asking what those services were for or what would happen if the funding disappeared.
Kearney, Deloitte and State of Flux independently identify the same 6% of organisations as elite. The research model that produces the number also explains why 94% can't close it.
BCG calls the CFO confidence gap a procurement credibility problem. It isn't. It's a definition problem that procurement created and then blamed finance for not understanding.