So the question is really how can you quantify the amount of your unreacted reactant in your product stream. The easiest way would be to include an internal standard at a known concentration. 

Without doing that:

Conversion is defined such that: Fout=Fin(1-X) -> Fout/Fin=(1-X)

So if you can get the ratio of your peak areas for your molecule of interest in the feed and product you should be able to get the conversion. 

Hope this helps

It possible to calculate the conversion without using internal solvents.You can remove the sample time to time take GCMC chromatogram. find our the area under peak of starting material & product .  Take ratio of the product  to the reactant x100= percentage  conversion. However,  every time you haveo scan GCMS. 

If u r getting different peaks compared to control then scan it for ms.

A quick comment. You use the term "the conversion of a product," which is not correct. The definitions  of "conversion" by commenters are also wrong. 

The conversion is (Co-C)x100%/Co

The product yield based on initial material is (Cp/Co)x100%

The selectivity is (Yield)/(conversion)

The use of internal standard compensates the variations of the injected volumes and thus decreases the experimental error. The use of internal standard is not obligatory.   

If your product is known than its simple to compare the urea under curve of GC with the reference material accouting for the dilution factors. Reference material should be pure enough

% yield= area under curve of reaction mixture/ area under curve of ref mat* Dilution Factors* 100

I think it will help is the case is as define above

good luck

Thanks a lot for your valuable comments.

By :internal solvents" I presume that you mean "internal standards".

You do not necessarily have to use an internal standard - whether it will improve your results depends on what you are analysing, how your prepare the samples, and your instrument method.

You do have to prepare a calibration for your target analyte so that you can quantitfy it - an internal standard is sometimes helpful for calibration, but it may not be neccessary.