I  think by using GC MS you predict the compounds of you samples, by comparing the (EI)MS spectra to the free on line MS data bases such as MassBank, NIST, or others. I think your GCMS has also MS Library. For confirming the prediction of the data bases than you can compare their MS to the MS of the standards. Using TLC, the identical Rf does not means identical compound.

For an internal standard you need a compound or compounds which are similar to the compounds present in your samples but something that does not appear in your samples (as that would make the results impossible to interpret accurately).  I agree with Gunawan that you need to do a preliminary analysis of your samples (maybe not all but a selection) to see what type of compounds are normally present.  This then allows you to choose an appropriate internal standard.

Normally when you do FAMEs by GC it is use FID detector. If you have the chance of using GC/MS normally it is use one methyl fatty acid deuterated (D3) as internal standard.

If you are using GC-FID I use heptadecanoic methyl ester, but first you have to see that it is not present in your matrix.

In TLC it is use one DIPA, you have to see which fit best for your sample. Disopropylamide of one specific fatty acid.

best

100% esterification does not occur! In addition the response of the detector to each individual fatty acid ester varies (unless the standard includes this molecule). Thus, fatty acid profiles are used for identification purposes. Check the USP specifications for corn, canola, and soybean oils.

Find out what is in your matrix 1st!