COEFFICIENT OF VARIATION :(

[SmileyScience]

What do you guys consider as acceptable cv when you analyse your data that is for example in triplicates (such as real time pcr triplicates)?
I was told that if cv is less than 25 = variability is ok

Please help

[Katy_Bobbles]

I was told it should be less than 5. A bit of a difference there but I suppose it may depend on the method used. I used spectrophotometry and the Biuret method to assess QC albumin levels in 20 samples. Hmm