Lynnerup & Ansell
Prepared for ECMC Foundation · Confidential
Lynnerup & Ansell
Prepared for ECMC Foundation · Confidential
What this is. An outside-in answer to the three research questions in ECMC Foundation's Postsecondary AI Funding Landscape RFP, assembled from public filings, funder grant databases, federal award records and secondary research. It was built without conducting a single interview.
How to read the plot. Each point is a topic. Its position on the horizontal axis is how much funder activity we could verify. Its position on the vertical axis is how strong the evidence of institutional need is. The upper left is where need is well evidenced and funders are absent. The lower right is where money is concentrated on claims the evidence does not yet support.
What it deliberately does not do. No invented figures. Where a number could not be sourced, the cell reads unsourced rather than carrying an estimate. Cash and in-kind are never combined. No individual at a peer foundation is named in an evaluative context. Treat this as the floor of what the commissioned research would establish, and not its ceiling.
Research Question 3
The RFP asks for these two elements mapped against each other, with areas of saturation and opportunity identified. This is that map. Click any topic for its evidence.
Research Question 2
The five domains named in the RFP, each rated for how much evidence actually stands behind claims made about it. None of the five rates strong.
Research Question 1
Sixty funders and vehicles, in three strata. Peer candidates are the reconstruction of ECMC's Center for Effective Philanthropy cohort, which is confidential to ECMC and selected by the commissioning funder; we intersected CEP's published roster with postsecondary and workforce fit, and it remains our assumption. AI-native covers technology and corporate funders a mission-defined cohort would never surface. Vehicles are pooled funds and federal programmes. Cash and in-kind are never combined, because most headline corporate figures are licence value rather than money.
| Funder | Total grants paid | Committed to AI | Type | Basis for the AI figure | CEP GPR |
|---|
Interrogate the evidence
Grounded strictly in the compiled evidence base. It cites the records it uses, and when the evidence does not support an answer it says so and names the interview that would resolve it. The refusals are the point.
The proposal
Submitted 31 August 2026 in response to the RFP released 15 July. Five months, November 2026 through March 2027, delivering the full scope of work: fifteen to twenty peer funder interviews, synthesis across the five institutional domains, and the two mapped against each other. The baseline below is what a weekend produced without interviews. The engagement is what converts it into knowledge.
| Phase | When | Deliverables and milestones |
|---|---|---|
| Mobilisation | Nov, wks 1–2 | Confirmed cohort and interview list. Agreed protocol. Instrument re-baselined to the confirmed cohort. |
| Data collection | Nov–Dec | Interview log with graded findings. Extended evidence base. Interim read-out to staff in mid-December. |
| Synthesis | Jan–Feb | Draft themes memo. Draft saturation and opportunity map. Working session with staff. |
| Deliverables | Feb–Mar | Full report. Executive summary. Visual framework. Staff presentation. Instrument and data transfer. |
| Budget line | Rate | Hours | Amount |
|---|---|---|---|
| Rasmus Lynnerup, Senior Partner and Engagement Lead | $450/hr | 140 | $63,000 |
| Charles Ansell, Senior Partner | $450/hr | 100 | $45,000 |
| Research Analyst | $150/hr | 120 | $18,000 |
| Instrument hosting, compute and model usage, 12 months | $12,000 | ||
| Travel and expenses | $12,000 | ||
| Total fixed investment | 360 | $150,000 |
Rates are fully loaded. There is no separate indirect or overhead charge and no subcontractor line. At standard rates this scope prices at $178,500; the total reflects a $28,500 reduction combining our non-profit discount with a further reduction recognising that the baseline research is already complete.
Twelve pages. Technical approach, work plan, stakeholder engagement, qualifications, line-item budget with optional add-ons, intellectual property disclosure, and references.
Download PDFVerified primary intelligence. Unstated theories of change, the budget reality behind the public number, what funders quietly stopped funding. Desk research cannot reach any of it.
A living instrument. The field will move between award in October and delivery in March. Twelve months of hosting and model usage are inside the fee, so what ECMC keeps continues to work.
The first AI learning agenda in the field. No peer funder has published one. Question 1(d) asks what they are; on present evidence, ECMC can author the answer.