After the Story: The Methods Are Open

If any of this felt like your shop, here is the part the novel was always walking you toward.

Everything Maya learned the hard way, nineteen days into a new job with most of a campaign on the line, is written down, in the open, and free. This time it’s method instead of a story. The problem this book dramatizes is the same problem a group of us have been quietly documenting for the sector, vendor by vendor, definition by definition, so that no one else has to learn it on a deadline with a board in the room.

It lives at fundcommons.org, and it’s meant to be put to work, not just read.

A few doors in, depending on where you’re standing.

If you don’t yet know how bad it is. Start with the self-assessment. It’s a short, honest diagnostic, the same questions Maya couldn’t answer in chapter one, and it tells you which rung of the ladder you’re actually on and what to fix first. You can take it before you tell anyone you’re worried. See the maturity self-assessment.

If you want to fix one real thing. The handbook isn’t organized as dry theory. It maps directly to the three maturity rungs (Clean, Connected, Predictive) that Maya climbs. If you are struggling with a specific operational headache—like reconciling campaign reports or cleansing a lapsed mailing list—you can jump straight to the relevant chapter of the novel to see the principles in action. You don’t have to tackle the whole system at once. Pick the rung and the chapter that matches your shop’s current pain point, and build a single, focused remedy. You can find worked templates, code adapters, and implementation partners in the ecosystem directory at fundcommons.org/ecosystem.

The shape Ruth kept pointing at. When Maya finally stops patching reports and fixes the shape of her data, so that a commitment is no longer confused with a transaction, a donor is one person to every query, and a fund means the same thing in finance and in advancement, that shape turns out to have a name. It’s the Advancement CRM Data Model: an open, vendor-neutral set of names for the handful of objects every fundraising database is really keeping, no matter what the vendor calls them on your screen. It belongs to no company, and you can map your own system to it without leaving it. It’s developed in the open at github.com/fundraising-commons/common-data-model. It’s young and still changing, which is the point of doing it in public.

All of it is openly licensed. Use it, copy it, adapt it for your shop, teach it to the next person who inherits a database nobody trusts. The whole reason it’s a commons and not a product is that a foundation owned by no vendor is the only kind that can actually close the gap this book is about instead of widening it.

What this builds on, and what it doesn’t replace. None of this is a vacuum, and it would be dishonest to pretend it were. The split the reconciliation bridge rests on — what a campaign has committed versus what it can recognize — is the line between CASE’s Global Reporting Standards, which is how the sector counts campaign production, and GAAP, which is how your auditor books revenue. The Commons doesn’t reinvent either; it just makes the two reconcile on one page. The definitions of a major gift, a solicitation, a contact, and the ethics of prospect research have homes too, at AASP and Apra, and Marcus’s accuracy is not permission is their line before it was his. What no standards body has shipped — the gap this book is actually about — is an open, vendor-neutral data model underneath all of it, with the worked practices to run it. That is the one thing the Commons is for. Naming what already exists is the point: you don’t have to start from scratch, and you don’t need anyone’s permission.

The tools are free. The scarce thing isn’t. You can build every floor in this book on software that costs nothing. A local analytical workspace — DuckDB — instead of a data warehouse you can’t afford. OpenRefine and Splink for the record-linkage that turned three Eleanors into one, so Reggie knew Walter becomes something a machine can do twice. Great Expectations or dbt tests to assert Nadia’s front-door rules instead of hoping for them. scikit-learn and pandas for the scoring and the out-of-time holdout that has to fail before you trust it. Metabase or Superset so every report reads one definition. And the ACDM to give all of it one set of names. License cost was never the wall for the lean shop. The wall is skill and time — the person who can drive the tools — and open source takes the price of the map and the tools to zero while leaving that wall exactly where it stands. So the Commons only closes the gap if it ships the practices: the maturity rungs, the worked examples, the templates a real person can follow on a Tuesday. Not just a schema. That’s why this is a handbook and a novel, and not a spec.

One honest word about the clock. The novel compresses, the way novels do. In a real shop each slice is a weeks-long job, the durable version is a staffed function that never entirely finishes, and nineteen days buys you a few good decisions and one defensible proof on a single question — not a clean file. Maya’s three weeks bought the bridge and the nerve to fund the rest. The rest was the work, and the work is the point.

You don’t have to do what Maya did. You don’t have to wait for the cold drop of being asked to defend a number you can’t. The methods are sitting right there, and they were built by hand, by people who care, long before any machine could stand on them.

Go build the foundation.

The author