Howard: a capacity and standing-charge result

West Mercia Energy's account, dated 16 December 2025, says The Howard Multi Academy Trust moved to its managed service and flexible procurement strategy in September 2024. WME reports £33,000 a year in standing and capacity-charge savings following a Maximum Import Capacity review and band reallocation. A named trust finance and operations officer describes capacity reductions at five sites.

Result status: provider-reported ongoing annual savings, supported by an attributed school testimonial. The result has not been independently audited here. The mechanism is capacity and standing-charge optimisation, not proof that a commodity tariff was £33,000 cheaper. Public source gaps include the full before/after bill model, service fees and independent verification. It illustrates that managed services can add value through an approved buying route too. Any capacity change at another school needs its own current and future load assessment.

United Learning: a four-year managed-contract claim

Inspired reports a £79,500 saving over the life of a four-year risk-managed electricity contract for United Learning, alongside a renewable electricity product and support with metering, billing and accruals. The account also describes flexible gas procurement and a framework relationship dating from 2014.

Result status: provider-reported saving over the contract term. The public page does not supply the detailed electricity comparator, meter scope, fee reconciliation or evidence establishing that the entire amount was realised. Its reference to 2018/19 prices concerns gas and should not silently become the electricity saving baseline. The framework relationship shows that adviser involvement and framework procurement can coexist; the page does not establish every procedural detail of this particular contract. The practical lesson is to ask for a whole-term cost and service comparison with its baseline stated.

Wellington: identified potential still needs verification

Zenergi's Wellington School page identifies potential reductions of 447,355 kWh and £31,720, following a two-day audit. It describes improvements to heating management, valves, lighting and energy awareness. The page also identifies £5,185 of annual savings associated with measures costing £2,000.

Result status: identified or estimated project potential. Although the headline uses stronger language, the body describes what the school could achieve. It does not provide a full post-installation measurement record proving that the totals were delivered. The earlier downloadable case also presents opportunities and estimates, with some differently grouped project figures. We therefore do not combine those totals or call them realised savings.

This is an energy-audit example, not a comparison proving that an outside-framework supply deal beats DfE. The next evidence would be implementation dates, costs and consumption adjusted for weather and operating conditions.

The official counterweight: DfE's historical benchmark

DfE's June 2026 analysis compared participating state-funded schools' 2024/25 bills with estimated costs for the same historical consumption under its government buying basket. Median potential differences were positive, but some incumbent arrangements were cheaper. This is important evidence to include when evaluating a TPI-supported alternative.

Result status: retrospective model, not a current quote or a guarantee. Schools volunteered, the sample covered a small part of the estate and the underlying bill and basket-price data were not published. It does not establish independent fee-paying school outcomes or an average causal benefit from using or avoiding a broker.

Use the cases in your own decision paper

These cases support a practical investigation, not a league table. They concern different dates, schools, services, baseline prices and cost mechanisms. Do not add separate sample medians or project estimates into a forecast for your school. Ask your adviser to identify which mechanism is relevant to your meters and produce the evidence needed to test it.

Our evidence standard asks five questions: Who reports the result? What exactly changed? What was the comparator? Is the amount forecast, recurring, recovered or actually paid? What fees, risks and missing data remain? A strong case answers those questions even when the answer is that a competing approved route delivered better value.