Guide
Offline Field Data Collection Checklist
Most field data problems are not discovered in the field. They are discovered weeks later, when someone tries to analyse what came back. This checklist is for working through before a team goes out.
Coverage and devices
Does the tool genuinely work with no connection at all?
Not "works on slow connections". No connection. Test it in aeroplane mode before you rely on it. Some tools appear to save and then silently discard the submission.
What happens if the app is closed or the device restarts mid-form?
Partially completed forms should survive. Ask, then test it.
How long can a device hold unsynced data?
If a team is out for a fortnight, the device needs to hold two weeks of submissions, photos included, without hitting a storage or retention limit.
Whose devices are these, and what happens to the data on them?
Personal phones holding unsynced community data is a governance problem worth naming before it happens, not after.
Consent and permissions
Is consent captured with the record, or tracked separately?
Consent recorded in a separate spreadsheet drifts out of sync with the data almost immediately. It is far more robust attached to the record itself.
Can consent be withdrawn later, and would you be able to act on it?
If someone withdraws, can you actually find and remove every record, photo and derived report? If the answer is "probably", the answer is no.
Do access restrictions travel with the data?
If certain records are restricted, that restriction needs to hold in the database, the dashboard and any export. Restrictions that exist only in the collection app are not restrictions.
Who has agreed to this, and at what level?
Individual participant consent, organisational approval and community or cultural authority are three different things. Know which ones your work requires before it starts.
Media and metadata
Do photos, audio and GPS stay bound to the record?
The classic failure: 300 photos in a camera roll and no reliable way to tell which record each belongs to.
Is location captured deliberately, or incidentally?
Photo EXIF data may embed coordinates you did not intend to collect. For culturally sensitive sites that matters. Decide, rather than discover.
Sync and integrity
How do you know a sync succeeded?
There should be a positive confirmation and a way to see what has not yet arrived. "It looked fine" is not a sync log.
What happens if two people edit the same record offline?
Conflict handling is where offline tools quietly differ. Silently keeping the last write can lose real work.
Can you tell a duplicate submission from a genuine repeat visit?
This needs to be designed into the form, not untangled afterwards.
Data quality
Is validation happening at the point of capture?
An out of range value caught on the device costs seconds to fix. Caught at analysis, it costs a return trip or a discarded record.
Are free text fields doing work that a structured field should do?
Free text is fast to design and expensive to analyse. Every open field is a decision to defer effort to your future self.
Has someone who will actually use the data reviewed the form?
Forms designed without the analyst reliably collect the wrong things.
Have you run a pilot with real staff in real conditions?
Run it before the full deployment, not after. A short pilot with real staff in real conditions surfaces the problems a feature comparison never will, and it is far cheaper to find them then than mid season.
If most of these are fine, you may not need us
If you only need to collect responses and read them back, a well configured off the shelf survey tool is often the right answer, and cheaper than a build. The case for building starts when that data has to flow into reporting, mapping, or systems you already run without manual re-entry. Our comparison page covers that trade-off honestly.
Planning a field project?
Talk it through before you commit.
If this checklist surfaced something awkward, that is worth a conversation, whether or not it turns into work for us.
Email hello@databytesai.com.au