# Stewardship Invests in Contributor Capacity, Not Just Cash The dominant corporate model for "supporting open source" is to write a check — a foundation donation, a FOSS fund, a sponsorship tier. That model addresses a real scarcity (money) but not the binding one. The binding scarcity in most critical open-source projects is contributor capacity: skilled human attention from people who understand the codebase well enough to triage, review, and maintain it. Money cannot be converted into that capacity on demand, because the bottleneck is not funding maintainers can hire against — it is the existence of trained contributors at all. Stewardship is the practice of investing in that capacity directly, and it is a categorically different intervention than funding. Bloomberg names the distinction in operational terms. Their "Sustaining Open Source" program runs time-bound mentorship cohorts where their own engineers — "including many who have never contributed to open source" — learn to contribute directly to a project under maintainer guidance, focused on "high-leverage maintenance work that maintainers rarely have time for, such as issue triage, tests, docs, small-to-medium fixes, examples, and tooling" (Source: [[3. Resources/Readwise/Articles/Sustaining OpenTelemetry Moving From Dependency Management to Stewardship]]). The framing is explicit: stewardship targets "the key missing ingredient: creating sustained contributor capacity for maintainers and projects." The contrast is with "dependency management" — track, scan, patch — which "mostly addresses what's visible." The real risk "lives below the surface": maintainer burnout, thin contributor pipelines, "critical projects that are 'everywhere' but owned by no one." After cohorts with pandas (via NumFOCUS) and a cross-industry run with NVIDIA, the consistent outcome was that "maintainers got meaningful relief on the operational load that has typically blocked long-term progress" — relief that the equivalent dollars could not have purchased, because there was no one to hire. George Chellapa's account from the NVIDIA OSPO sharpens why cash alone fails. His FOSS fund started at a deliberately small $40K — "as we experimented so that we could build the internal processes" — and the hardest parts were not the money but the operational machinery: aligning with finance on fiscal hosts, getting payments processed, qualifying projects (Source: [[3. Resources/Readwise/Articles/Setting Up an Open Source Fund – Key Learnings to Date - George Chellapa]]). He grounds the urgency in the same structural fact this note turns on: 96% of scanned codebases contain open source, firms "would need to spend 3.5 times more" without it, yet "most open-source projects are maintained by one or two maintainers and 60% of them identify themselves as unpaid hobbies[ts]." A fund can move money to those maintainers; it cannot create the second and third maintainer that would actually de-risk the project. That requires growing contributors — stewardship's job. The third source generalizes the pattern to the ecosystem level. The All Things Open report on the Linux Foundation Member Summit frames open source as "critical infrastructure, not just a development model or a cost strategy," and points to the Bloomberg+NVIDIA pandas collaboration as the model in practice — "more than 1,000 volunteer hours and over 70 improvements to pandas" in ten weeks — explicitly as what happens "when organizations move from isolated contributions to coordinated collaboration" (Source: [[3. Resources/Readwise/Articles/Open source is critical infrastructure, not just a development model We Love Open Source • All Things Open]]). The unit of value is contributor-hours coordinated into a project's maintenance backlog, not dollars deposited. A crucial, non-obvious corollary: stewardship targets *healthy* projects on purpose. Bloomberg chose OpenTelemetry precisely because it is "a healthy, well-governed project — and that's exactly the right moment to invest in its contributor pipeline." Stewardship is prevention, not rescue. Funding tends to flow to projects already in crisis (the post-incident sponsorship surge); capacity investment is most effective applied before "a project's human bandwidth can erode" to the breaking point. This inverts the usual triage instinct and is the single hardest part of the model to sell internally, because the payoff is an outage that never happens. ## Why This Matters For **OSPOs and corporate open-source strategy**, the implication is that a funding line item and a contributor-capacity program are not substitutes, and a strategy with only the first is incomplete. The right question is not "how much do we donate?" but "how many of our engineers are becoming contributors to the projects we depend on, and to which projects?" An org running a six-figure FOSS fund with zero structured contribution is buying reputational cover, not resilience. For **maintainers and foundations**, stewardship reframes what to ask corporate partners for. The high-value ask is not (only) money but mentored contributor time aimed at the maintenance backlog — triage, tests, docs — the unglamorous work that money rarely attracts and that burns maintainers out. Foundations that can broker and structure cohorts (NumFOCUS, CNCF) become capacity-routing infrastructure, not just fiscal hosts. For **sustainability discourse**, this note narrows a vague term. "Open source sustainability" is often discussed as a funding problem (Tidelift, GitHub Sponsors, the Maker-Taker debate). The funding frame is real but incomplete: it solves for maintainer compensation, not maintainer multiplication. The projects most at risk are not underfunded so much as under-staffed — one or two people deep — and that risk is unaddressable by payment alone. ## Counterpoints to Hold - Money and capacity are not cleanly separable: a funded maintainer can reduce hours at a day job and mentor new contributors, converting cash into capacity over time. The claim is not that funding is useless but that the conversion is slow, indirect, and frequently doesn't happen because the maintainer uses the money to survive, not to train successors. - Mentorship cohorts have real costs the funding model avoids: they consume senior-engineer and maintainer time (the scarce resource itself) to onboard newcomers. A poorly run cohort can *subtract* capacity by taxing maintainers with low-quality contributions. Bloomberg mitigates this with structured onboarding and office hours, but the failure mode is real and is why this works better on healthy projects. - For some projects, the binding constraint genuinely is money (a solo maintainer who would happily continue if they could afford the time). There, a fund is the correct first intervention. Stewardship's claim is about the *typical critical-infrastructure* project (deep, widely-depended-on, 1–2 maintainers), not every project. - Corporate contributor programs carry capture risk: engineers from one vendor steering a project toward that vendor's needs. Vendor-neutral foundation brokering (CNCF, NumFOCUS) is the structural guard, and the cross-industry (Bloomberg+NVIDIA) framing is partly a response to this. <!-- atomicity: 1 idea ✓ --> ## Links - [[3. Resources/Permanent Notes/Zet · Technical excellence alone cannot overcome economic misalignment]] — applies-in: aligned funding has to build the people and maintenance capacity that technically sound shared infrastructure still depends on - [[3. Resources/Permanent Notes/Intent Written on the Outside Compounds While Re-Derived Intent Is Debt]] — applies-in: durable written intent transfers context to future contributors and compounds the capacity of every later run. - [[3. Resources/Permanent Notes/A Project Is the Community Not the Code]] — extends; investing in contributor capacity follows from treating the community, not the code artifact, as the durable project asset. - [[3. Resources/Permanent Notes/Maintainer Attention Is the Scarce Resource in Open Source]] — this note is the corollary: if attention is the scarce resource, capacity-building (not funding) is the intervention that targets it - [[3. Resources/Permanent Notes/Maintenance Is Most of the Work but Gets No Prestige]] — stewardship's focus on triage/tests/docs is precisely the unglamorous maintenance work that prestige systems ignore - [[3. Resources/Permanent Notes/The Maker-Taker Frame Confuses Two Different Scarcities]] — directly adjacent: the maker-taker debate is a funding-frame argument; this note argues the more binding scarcity is capacity, which maker-taker doesn't address - [[3. Resources/Permanent Notes/The Inverse Commons]] — the abundance/scarcity inversion that explains why the bottleneck is people, not code - [[3. Resources/Permanent Notes/Casual Contributors Are Users Not Community Members]] — the capacity-pipeline question (turning users into contributors) is exactly what stewardship cohorts operationalize - [[3. Resources/Permanent Notes/A Contribution Pipeline Needs a Defined Next Step After Training]] — because applies-in; contributor capacity is not retained when a training program ends without a specific post-training path. - [[3. Resources/Permanent Notes/Public Credit Mechanisms Are the Visible Skeleton of Open Source Social Contracts]] — credit is the retention mechanism that turns a one-time cohort contributor into a sustained one - [[3. Resources/Literature Notes/The Key Steps to Building a Community Mentorship Program]] — applies-in: mission, matching, expectations, resources, and feedback are the program mechanics that determine whether mentorship creates net contributor capacity - [[3. Resources/Permanent Notes/Curated Mastery Conferences Serve Practitioners That Volume Conferences Cannot]] — applies-in: paying expert speakers treats practitioner knowledge as capacity worth investing in rather than volunteer labor to extract. - [[3. Resources/Permanent Notes/Zet · Enterprise open-source participation needs policy and training scaffolding]] — applies-in: enterprise policy and training invest directly in employees' capacity to participate upstream. - [[3. Resources/Literature Notes/Lit · Pay WordCamp organizers and speakers]] — applies-in: paying organizers and speakers treats their knowledge and time as contributor capacity worth sustaining. - [[3. Resources/Literature Notes/Open Source and United Nations Goals]] — extends: decent open-source work requires building contributor capacity, not merely aligning public language with sustainability goals. - [[3. Resources/Literature Notes/Open Source in Enterprise Development]] — applies-in: enterprise adoption becomes durable when policy, training, and mentorship build employees' ability to contribute upstream. - [[3. Resources/Literature Notes/The Feasibility and Governance Concerns Behind 5% Contribution to WordPress Core]] — contrasts: capacity-building support offers a better stewardship model than treating five percent as a flat staffing quota. - [[3. Resources/MOCs/Events/WordCamp Europe 2026 MOC]] — applies-in: the MOC lists this note under 🌐 Open Web, Open Source & the Commons. ## Open Questions - What is the measured retention curve of cohort contributors? How many keep contributing six months out, and what predicts it? Bloomberg promised a post-cohort report; the data isn't public yet. - What is the senior-engineer/maintainer cost per unit of capacity created? Stewardship spends the scarce resource (attention) to make more of it — under what conditions is that net-positive vs. net-negative for the maintainer? - Does the model work for *unhealthy* projects, or only healthy ones? If only healthy, what fixes already-eroding critical projects — and is that just funding plus emergency hiring? - Is there a measurable substitution rate between dollars and contributor-hours, and does it vary by project maturity? The natural experiments (funded-only vs. capacity-only vs. both) exist across the ecosystem; the comparison doesn't. ## Source - [[3. Resources/Readwise/Articles/Sustaining OpenTelemetry Moving From Dependency Management to Stewardship]] — Tech at Bloomberg (Mar 2026): the canonical statement of stewardship vs. dependency management; "sustained contributor capacity" as the missing ingredient; healthy-project-on-purpose - [[3. Resources/Readwise/Articles/Setting Up an Open Source Fund – Key Learnings to Date - George Chellapa]] — George Chellapa, NVIDIA OSPO: the funding-side practitioner view; the 96% / 3.5x / 60%-unpaid-hobbyist scarcity data; why the hard part is operational, not financial - [[3. Resources/Readwise/Articles/Open source is critical infrastructure, not just a development model We Love Open Source • All Things Open]] — All Things Open (June 2026): the ecosystem-level "critical infrastructure" framing; the pandas 1,000-hours/70-fixes result as coordinated capacity in action - See also: [[Open Source Governance MOC]] for the governance context and [[AI Cost & Tokenomics MOC]] for the AI "review tax" that raises maintainer load (the why-now)