How we'd measure across the three bets and their flagship partners: which evidence tool, for which partner, when. The green dot (●) marks evaluation/measurement already underway — including the World Bank RCT in Lima and the Education Reports / Clio analytics.
North star · US
Halve the income-based NAEP reading & math gap in 2–3 focus states (starting Detroit, MI).
North star · Global
Cut learning poverty from 80% → 40% in 2–3 focus countries (Rwanda, India, Ghana, Peru).
Showing what's underway / starting now — step through the horizon to sequence the roadmap, or hit All.
How to read this. Columns are Education's flagship partners, grouped under the three bets; rows are measurement stages — Adoption & use → Delivery → Impact (proximal learning outcomes, then the distal north-star gap). Chips are the evidence tools, each carrying a phase number (1 Now → 4 6mo+) for when that tool, for that partner comes online. A green dot (●) marks work already underway: the World Bank RCT in Lima (the marquee causal study, via Project Primer), plus Clio usage analytics behind the published Education Reports. North-star metrics (NAEP gap for US partners, learning poverty for Global) are the distal endpoints, measured against baseline / the World Bank's 2030 trajectory — long-horizon by design. Bet 3 partners (Teach for All, Playlab) are R&D, so their "learning outcome" is the hypothesis itself (AI-fluency, redesigned-school performance) and they carry no near-term north-star claim. MagicSchool is TBD, so its instruments sit later. Priority shading is intentionally omitted here — unlike the portfolio roadmap there's no per-partner decision-relevance grid for education yet; the hero on this view is sequencing. Hover any cell for detail.