SCRNA-seq experiment reads

Question

[kanwarjag]

I am planning a 10X single cell RNAseq library preparation and sequencing in Nova seq. For human genome to look at standard differential expression and cell types, how many reads are required / sample. Is there any standard (not rare) rule of thumb how many cells should be sequenced. I have read 10X information @ https://kb.10xgenomics.com/hc/en-us/articles/115002022743-What-is-the-recommended-sequencing-depth-for-Single-Cell-3-and-5-Gene-Expression-Libraries- ta suggest 10K reads per cell. Do we do estimation per sample? Thanks

[ATpoint]

What is unclear after reading the official recommendation?