How do you prioritise SEO work when you have limited time and resources?

Give a framework, because SEO generates infinite possible tasks and the skill is choosing.

Prioritise on impact against effort:

• Fix anything blocking indexing first. A page Google cannot crawl or index earns nothing regardless of how good it is. This always outranks optimisation work.
• Then pages already close to ranking. A page at position 8-15 with good impressions needs far less work to reach page one than a new page needs to rank at all. Search Console makes these easy to find, and it is the highest-return routine task in SEO.
• Then high-commercial-intent pages. Ranking for a term that converts is worth more than ranking for one with ten times the volume and no intent.
• Content updates before new content. Refreshing pages that have decayed is usually cheaper and faster than earning authority for something new.
• Structural work — internal linking, site architecture, templates — because it compounds across many pages at once.

Note: Mention that you factor in implementation reality. A technically ideal fix requiring six months of developer time may be worth less than three smaller changes you can ship next week. Working within constraints is what makes an SEO effective in-house.