EXPOSED
The production half of this job — drafting appeal letters and case statements, writing grant proposals, segmenting donor databases, building prospect research briefs, assembling board reports and campaign dashboards — is already well within reach of current AI, and small shops are the first to substitute. The surviving half is the part no model can do: sitting across a table from a donor capable of a seven-figure gift, reading the room, deciding the ask amount and timing, and being the accountable face when a campaign misses its number. There is no licensure here (CFRE is voluntary), so the moat is relationship and outcome ownership, not regulation.
Mixed — a routine tier and a judgment tier. Roughly half the week — appeal copy, grant narratives, wealth-screening summaries, RFM segmentation in Raiser's Edge or Salesforce NPSP, gift-acknowledgment templates, campaign pacing reports — is generative or query work a model does in minutes, which is why this lands at 9 rather than 14; what holds it above 6 is the solicitation cycle itself, cultivation visits, board-member coaching, and volunteer committee management that has to be done by a person.
Some physical or field component. A 7 reflects the travel, not the labor: donor lunches, cultivation dinners, campus and program-site tours, gala and groundbreaking events, and regional prospect trips put you in rooms and cars, but nothing about the physical setting is uncontrolled or skill-bearing — the work itself is talk, and the rest of the week is CRM and Zoom.
No licence, no signature requirement. CFRE and ACFRE are voluntary credentials no employer is legally required to demand, and nothing bars an unlicensed person from running a capital campaign; the compliance exposure that exists — IRS substantiation letters, state charitable solicitation registrations, Form 990 Schedule G — attaches to the organization and its officers, not to your personal certificate, which is why this sits at 3.
The human relationship is the product. Major-gift portfolios are named relationships built over three to seven years, and when a development director leaves, the top-twenty donors' giving frequently follows or stalls until re-cultivated — the donor gives because of who asked, which is why 16 rather than 12; it stops short of the top because direct mail, digital, and grant revenue flow to the institution regardless of who signs the letter.
Exists to be accountable for ambiguous calls. You set the ask amount and the moment for a prospect with one shot at a seven-figure commitment, decide whether a gift with restrictive naming or reputational baggage should be accepted or refused, and set the campaign goal and feasibility threshold the board will be held to — calls made on incomplete wealth data with no procedure to fall back on, which puts this at 14 even though the board formally votes.
Has AI actually changed your work?