Digitisation projects fail quietly. The scanning finishes, the boxes go into storage, and eighteen months later staff still ask the registry clerk to find things — because the scanned files are named by batch number and nobody can search them.
Four decisions, made before scanning begins, prevent this.
Decide what not to scan
A retention schedule usually reveals that a substantial share of a registry is past its required life. Scanning it costs money and then costs storage. Establish what may lawfully be destroyed, get that decision signed by someone with authority, and shrink the project before it starts.
Decide the index fields
A scanned page is a picture until something describes it. Agree the handful of fields by which people will actually search — client name, reference number, document type, date, department — and capture them at the point of scanning. Retrofitting an index later costs more than the scanning did.
Decide which copy is the record
After digitisation, the digital file must be the authoritative record and the paper either destroyed or archived as a non-authoritative copy. If both remain in use, staff will keep updating paper, and within a year the digital set is out of date and distrusted. This decision needs to be written down and communicated, not assumed.
Decide who may see what
Paper access is controlled by a locked room. Digital access is controlled by permissions, and if nobody sets them, the default is often everybody. Map document classes to roles before loading anything, so the migration applies the access model rather than importing an open one.
Then run it properly
With those four decisions made, the mechanics are straightforward: scanning standards and resolution, quality checks on a sample of each batch, indexing verification, and a destruction certificate for what was disposed of. What makes the difference is that the decisions came first.
An information audit produces all four in a few weeks and is useful even if digitisation is postponed.