Movement-Owned, Self-Hosted, or Off-the-Shelf

Beneath every tool-selection spreadsheet — the feature matrix, the pricing tiers, the integration list — sit three questions the spreadsheet never asks. Who holds your data? Who profits from your use? And what happens to both answers on the day the vendor is acquired? For campaigns, unions and community organisations, these are not procurement hygiene. They are questions about power, asked of the software that increasingly holds the movement's memory.

Three ways to hold a tool

Every option in the catalogue resolves to one of three ownership shapes. Off-the-shelf commercial software is the default: polished, well-resourced, cheap to start and supported at scale — with your data resident under terms you did not write and cannot renegotiate, subject to change with thirty days' notice.

Movement-owned tools are built by and for the sector. The good ones encode real organising knowledge, pledge not to train models on your content, and answer to people who share your goals. They are also, structurally, small companies: thinner roadmaps, fewer engineers, and a founder whose eventual exit options are a fact of life rather than a character flaw.

Self-hosted open source is the third shape: the automation platform or model runner on your own server, where the data never leaves the building. It offers the strongest answer to all three questions and charges for it in kind. Control is not a setting but a chore, paid for in a technical staff member's hours, security patches applied on time, and backups that are tested rather than assumed.

The acquisition test

The scenario worth planning for is not vendor collapse but vendor success. Within recent memory, two well-regarded movement tools were acquired within a year of each other — both continuing to operate, both now answerable to owners with broader commercial ambitions than the sector they grew up serving. No scandal attended either sale. That is the point: acquisition is the ordinary life cycle of venture-backed software, and a tool built for movements is not exempt from it.

So run the test before adoption, not after the press release. If this vendor is bought tomorrow, what travels with it? Your supporter records, your message history, your workflows and your team's muscle memory all sit inside the product, and a member list can outlive its protections when the entity holding it changes hands. The practical disciplines follow directly: ask about export before you import, read the data-processing terms for their change-of-control clauses, and prefer tools where leaving is a day's work rather than a quarter's. Exit cost, not sticker price, is the number that matters in year three.

What Australian organisations owe

None of this is merely prudential. An ACNC-registered charity should expect its handling of supporter and beneficiary data to be judged against the Privacy Act's principles — collect what you need, secure what you hold, be straight about why — whatever the technical reach of the exemptions, and the reform direction points toward fewer exemptions, not more. The ACNC's governance standards reach here too: a board's duty of care extends to knowing where the organisation's data sits and under whose law, and a committee that cannot answer those questions is not governing them.

Unions carry the sharpest version. A member list is leverage-sensitive in a way a newsletter list is not, and the acquisition test acquires teeth: member data inside a vendor later purchased by a firm that also sells services to employers is not a paranoid hypothetical but a due-diligence question with a place on the checklist. Even for organisations whose DGR status makes them attractive to trusts and foundations, the data questions now appear in funder due diligence — stewardship of supporter information has quietly become part of what grant-makers mean by good governance.

When the overhead is worth it

The honest answer to the build-borrow-buy question is a matrix, not a manifesto. Weigh the sensitivity of the data flowing through the tool against the cost of losing it. Generic drafting and research sit safely in off-the-shelf software at around A$30 a seat each month, provided sensitive content stays out of the prompts. Anything touching supporter or member records deserves either a movement-owned tool with contractual data protections or commercial software on a proper organisational tier — never a consumer account with the treasurer's password in a shared document.

Self-hosting earns its overhead at the top of the sensitivity curve — union membership data, casework records, anything whose exposure would harm the people it describes — and only when the organisation can genuinely staff it. An unpatched server run by a volunteer who left in March is worse than a well-run vendor; sovereignty without maintenance is just risk with better politics. Ownership, in the end, is not a purchase but a practice: a question the organisation asks itself annually, while leaving remains cheap enough to be a real answer.