Spreadsheets are reliant on software that could become obsolete. When this materialises, the file must be converted to an alternative format and, as a consequence, a loss of information could occur. It is therefore important to define what is significant and should be preserved. This study develops a framework on how to assess which properties stakeholders deem to be significant. Statistical analyses combined with a Spreadsheet Complexity Analyser show that objective criteria of a spreadsheet are strong predictors of significant properties. This study also provides guidance to archives for best practices concerning stakeholder analysis in the context of significant properties.