Right now, your formula counts the number of text values in column M.
That is not a robust approach because column M contains only five text values, but columns S and T have many more values.
If you don't know which column may have the most number of entries, you can introduce a helper cell in each column that counts the number of entries below. I suggest you insert a new row 2. In column M, for example, put a formula in M2
=counta($M$3:M$99999)
Copy that formula across to column T.
Next you can evaluate which of the columns has the largest number
=max(M2:T2)
This can be plugged into your original formula like this:
=OFFSET(Sheet1!$M$8,0,0,max(M2:T2),COUNTA(Sheet1!$1:$1))
So now, instead of just looking at how many rows are in column M, the formula uses the maximum number of rows in the columns M to S.
You can now hide row 2 if it upsets your worksheet design.
Edit: the mere count of text values with CountA will ignore blank cells and will return incorrect results. You really need a formula to find the row number of the last populated cell in each column.
This should really be a new question, but here goes
If the column has number values you can use
=MATCH(99^99,B5:B999,1)
If the column has text values you can use
=MATCH("zzz",C5:C999,1)
Adjust your ranges accordingly.