Under the lens of next generation sequencing technologies, the transcriptional profile of a single cell in its microenvironment and interactions can be analyzed by sequencing their RNA. In order to do it, the RNA must be extracted, converted to complementary DNA (cDNA), amplified and quantified. cDNA synthesis and amplification is done by the reverse transcription polymerase chain reaction (RT-PCR) using the RNA as template. RT-PCR tenet is that even though the complete process is exponential all the proportions are conserved, this process only helps to increase the accuracy of quantification by increasing the system scale. Therefore, the ratio between molecular counts of two genes must be the same before and after RT-PCR. However, the process is susceptible to technical errors and stochastic biochemical fluctuations. Particularly, quantification is affected for genes in lower count before RT- PCR, which leads to a zero count imputation for that gene. This non-error free process introduces a bias in data that reflect a particular condition. A lot of working has been done making statistical corrections by fitting error models to filter the data. Nevertheless, we postulate a new standpoint, the RT-PCR as a set of chemical reactions susceptible to environmental fluctuations; therefore, we used a stochastic chemical kinetics formulation to model RT-PCR process. We solved the system by applying the Gillespie's Algorithm. Our findings set that RNA to cDNA synthesis process is more prone to fail which is contrary to actual belief. These results demonstrate that RT-PCR is subject to large, intrinsically random fluctuations and raise questions on how to reduce the impact and number of errors in order to do not modify and properly interpreter the cellular reality.