Refreshing existing content usually returns more per hour than writing new content, and almost no programme allocates budget to it deliberately.
Find the pages that are declining, not the ones that are old
Age is a poor selector. Sort by traffic trend over twelve months and the candidates appear: pages that ranked well and are slipping. Those have demonstrated demand and existing authority, which is most of the work already done.
Update in place and keep the URL
Republishing as a new post discards the links and history that made the page valuable. Update the existing URL, revise the date honestly, and let the accumulated equity carry the improved version.
Decide what a refresh actually means
Changing the year in the title is not a refresh. Genuine updates replace outdated claims, add what has changed, remove what no longer applies, and improve the sections where competitors now answer better. Establish that standard or the programme becomes date-swapping.
Budget it as a fixed share
Allocating something like a quarter of content capacity to refreshing existing pages is a defensible split. Without a standing allocation it never happens, because new content always feels more like progress in a planning meeting.