Personally i would first examine the data to see if it is feasible.
For this you will probably need a microbiologist which i hope you possess.
IMO micro.data of the type you refer is likely to show considerable variation simply on account of sampling / testing factors.
i agree with the previous poster's caveats however if there is a substantial history of good correlation of seller/receiver data, it gives some optimism that the sampling/testing procedures are comparable.
IMEX it is beneficial to establish a timeline of seller/buyer (hopefully only one) data so as to know the achievable degree of similarity and thereby evaluate whether the complaint is within the usual range of correlation or not. you may be looking for a needle in a haystack.