Personally I won't not continue with the analysis using Excel. As in the wet lab, you must follow a procedure, which needs to be tracked, reproducible and readable: with Excel this is not possible, even if you're so crazy to script all the procedures using VBA. Here's the most comprehensive, yet simple for beginners, walkthrough analysis of RNAseq data analysis using R and the library DESeq2 https://master.bioconductor.org/packages/release/workflows/vignettes/rnaseqGene/inst/doc/rnaseqGene.html

Just to answer the question, I'll add a column with the function =COUNTIF($START:$END, 0), where $START is the first sample column (ex. B1) and $END is the last sample column (ex. G1), which returns the amount of columns with 0s. Then I'll filter the matrix using this column.