Integrations and your data
Where your data comes from, where it can go, and the habits that stop a campaign's most valuable asset turning into six inconsistent spreadsheets.
What is already there
You do not import Australia. The civic data is built in from day one:
- G-NAF addresses – the national address file, so the doors are real rather than guessed.
- ASGS statistical geography – the Bureau's standard framework, so your areas line up with how the country is actually measured.
- Electoral divisions at every level – federal, state and local.
- Politicians, policies and referendum data – the political context in the same place as everything else.
- Demographics – who lives where, not just where they live.
This is the groundwork most campaigns burn a fortnight assembling. Knowing it is there stops you rebuilding it.
What you bring
Your own list, and whatever comes from the tools you already run. Before connecting anything, decide one thing: which system is the source of truth for a person's contact details?
Pick one. Every integration problem that is not a bug is two systems both believing they are authoritative, quietly overwriting each other.
Sync direction is a decision, not a setting
For each connected source, write down:
- What flows in – new supporters, event RSVPs, petition signers.
- What flows out, if anything – usually less than people expect.
- What happens on conflict – if both sides changed a phone number, which wins.
If you cannot answer the third question, you do not have an integration, you have a race.
Import discipline
The rules that keep a list clean are boring and they work:
- One row per person. Deduplicate before upload.
- One phone format, applied consistently.
- A source column on every record. Where this person came from. You will want it constantly and cannot reconstruct it later.
- A date column. When they entered the list. Recency is the single best predictor of whether they will respond.
Check after every import: row count matches, ten random records look right, expected merges happened. See Audiences and segments.
Getting data out
Results export straight to CSV for the things that live outside the platform – a coalition partner, a mailing house, a scrutineering plan, your own analysis.
Treat every export as a copy of your database that you no longer control. That is not a reason to avoid exporting; it is a reason to be deliberate:
- Export the columns you need, not everything.
- Note who exported what and why.
- Delete the file when the task is done.
- Keep export rights with organisers. See Team, roles and access.
Publishing data on purpose
Polling views can be made public and embedded – dropped into a briefing, a website or a partner's page without exporting screenshots that go stale the moment the numbers move.
This is a genuinely useful campaigning asset: a live, credible picture of the electorate that supporters, media and coalition partners can see for themselves. Just be deliberate about what you make public, and remember that a public view stays public until you change it.
Data hygiene on a schedule
Once a quarter, thirty minutes:
- How much of the list has never responded to anything?
- How many records are missing a source?
- Which integrations have not synced recently, and does anyone still need them?
- Who has export rights, and should they still?
A list nobody prunes becomes a list nobody trusts, and a list nobody trusts gets replaced by a spreadsheet on somebody's laptop – which is how campaigns end up with six versions of the truth.