# Fundraising

> Raise money — from investors or from donors — by building the conditions where capital wants to come to you, then running the ask as a disciplined process. Use when the user is raising a round or a campaign, writing or critiquing a pitch deck or case for support, deciding whether to raise at all, targeting investors or donors, structuring a lead gift or lead investor, planning a capital campaign, improving donor retention, designing a matching challenge, handling investor or donor objections, or asking what actually makes people fund things. Covers both branches: startup/venture and nonprofit/philanthropy. Every technique is graded by the strength of its evidence — field experiments and large datasets separated from founder folklore. Not legal or tax advice.

- Skill: `dy/fundraising` (Agent Skill, multi-file: 3 files)
- Install (CLI): `npx skillmds@latest add dy/fundraising`
- Raw SKILL.md: https://api.skillmd.com/api/skills/dy/fundraising/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Marketing & Growth
- Author: DY (https://skillmd.com/u/dy)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/dy/fundraising

---


# Fundraising

Money follows conviction that already exists. Your job is to build that conviction before you ask, and then make the ask so easy to say yes to that saying no takes more effort.

**The governing finding.** Panas interviewed donors who had given a million dollars or more and asked what actually triggered the gift. Every one of them already believed in the mission before anyone asked. *The ask does not create belief; it converts belief into a transaction.* This holds identically on the venture side — the deck does not create investor conviction, it gives conviction a form to sign. Almost every fundraising failure is an attempt to do conviction-building and converting in the same meeting.

**The consequence — the ideal raise has no pitch.** Not because pitching is beneath you, but because a pitch is what you need when the evidence cannot speak. Work backward from that: what would make the ask a formality? Usually it is traction a stranger can verify, a first committed backer who has already staked their own name, and a constraint that is genuinely real. Build those and money starts arriving with less argument. Fabricate them and you have committed fraud with extra steps.

## The physics of allocation (both branches)

Five mechanisms decide whether capital moves. Four are well-evidenced. The fifth is mostly folklore, and knowing which is which is most of the value here.

1. **A credible first commitment is the strongest lever there is.** This is the best-evidenced finding in the whole field. List & Lucking-Reiley ran a real field experiment on a university capital campaign (*Journal of Political Economy*, 2002) varying how much announced seed money a campaign had: raising the seed share increased both participation *and* total dollars, and did so **more than linearly**. The venture analog is quantified too — CB Insights found that when a seed lead declines to follow on, the odds of closing a Series A fall to roughly 27%. Same physics in both branches: **secure the top before you go wide.** This is the empirical backbone under the capital campaign's quiet phase and under "get the lead investor before opening the round." If you take one thing from this skill, take this.
2. **Real signal beats narrative.** Early money responds to verifiable quality: traction and growth rate, preparation, network. Later money herds on early money. But the herd only forms if the first wave was signal-driven, so there is no shortcut that starts at the herd.
3. **Social proof works on the undecided middle, and has a ceiling.** Shang & Croson (*Economic Journal*, 2009) raised average public-radio gifts substantially by naming what a peer gave — but framing at the 99th percentile weakened the effect, because an unreachable comparison stops being a reference point. Frey & Meier (*AER*, 2004) found peer information moves the previously-indifferent, not the committed and not the never-givers. So the anchor should be **aspirational but reachable**, aimed at people who haven't decided.
4. **Warm introductions genuinely outperform cold outreach** — but every circulating multiplier ("13x", "10-20x") traces to single-firm funnels or vendor data, never a controlled comparison, and the populations differ systematically. Use the direction, discard the numbers.
5. **Scarcity is real only when it is structurally true.** A fixed round size, a closing date, a matching deadline that actually expires — those are constraints, and they move people. The "build urgency / manufacture FOMO" advice genre is unfalsified practitioner content layered on top of that real constraint. Perform urgency you don't have and you are running the same play as a resetting countdown timer, with a securities lawyer attached.

## The procedure

1. **Decide whether to raise at all.** Compute default alive or default dead (Graham): does current growth reach profitability before the cash runs out? Raising is not a milestone, it is a financing decision with a permanent governance cost. On the nonprofit side, the parallel question is whether you have the stewardship capacity to keep the donors you're about to acquire — acquiring donors you will lose is a net loss.
2. **Build the conviction before the ask.** Cultivation, evidence, proximity. The venture form is traction and warm relationships built months ahead; the philanthropic form is the moves-management cycle. Both are slow and neither is skippable. → [references/venture.md](references/venture.md) · [references/philanthropy.md](references/philanthropy.md)
3. **Secure the first credible commitment.** The lead investor, or the lead/challenge gift. Do not go wide before this. See mechanism 1 above.
4. **Sequence the rest around it.** Quiet phase before public phase. Soft commitments collected greedily, then selectivity once momentum is real (Graham's two phases).
5. **Make the ask.** Specific amount, specific use, specific person, in person where the sum justifies it. → the branch reference.
6. **Steward, and report before you ask again.** The single highest-leverage neglected step, and the one with the ugliest data behind it: roughly four out of five first-time donors never give a second time. On the venture side the equivalent is the investor update you send when the news is bad.

## Where to read

| The task | Open |
|---|---|
| Raising from investors, angels, VCs; pitch decks; term literacy; round mechanics | [references/venture.md](references/venture.md) |
| Donors, major gifts, capital campaigns, retention, matching challenges, grants | [references/philanthropy.md](references/philanthropy.md) |
| Persuading the *allocator* once you have their attention — proof, objections, framing | the `marketing` skill ([references/persuasion.md](../marketing/references/persuasion.md), [references/content.md](../marketing/references/content.md)) |
| Earning the third-party credibility that makes money arrive unbidden | the `public-relations` skill |

## The laws

- **Cultivate before you convert.** The ask is the last five percent of the work.
- **Never ask for money from someone who doesn't yet believe.** You will get a no *and* burn the relationship you could have built.
- **Secure the lead before you go wide.** Evidenced in both branches. Ignoring it is the most common self-inflicted failure.
- **Specific beats large.** A named amount for a named purpose outperforms an open-ended appeal, in both branches.
- **Report on the last gift before requesting the next.** Retention is cheaper than acquisition by an enormous margin and almost nobody does it.
- **Scarcity must be true.** If the constraint is real, state it plainly. If it isn't, you have nothing to say and should say nothing.
- **Discount every success story, including the famous ones.** Nearly the entire startup fundraising canon is written by people describing what worked for them, with no control group of founders who did the same and failed. Treat named rules as heuristics with no data behind them.

## Ethics (non-negotiable)

Manufactured inevitability is fraud. Fake oversubscription, invented traction, a "lead investor" who hasn't committed, a matching gift that doesn't exist, an impact number you cannot document — each of these is a material misrepresentation made to induce someone to part with money, and on the venture side that is securities fraud regardless of how the deck is worded. The mechanisms in this skill work *because* the signals they rely on are costly and true; counterfeiting the signal destroys the thing that made it work and takes the relationship with it.

This skill is not legal, securities, or tax advice. Term sheets, safes, cap tables, solicitation registration and charitable-deduction questions go to a lawyer.