Chemistry is an extremely powerful tool to estimate the duration of the prestellar phase, a key ingredient to distinguish between a slow or a fast path towards the formation of stars. In this talk I will focus on the potential of the relative abundance of o-H2D+ and N2D+ as an indicator of evolution. We have recently explained the opposite behaviours of the abundance of these species with time with simple considerations on the chemical formation paths, depletion of heavy elements, and evaporation from the grains. Our unique 3D MHD simulations, coupled with chemistry, take us one step further than the simple relative timeline determined by the ratio, allowing to follow its evolution with time. I will then discuss the independent estimate 
from the direct measurement of the ortho-to-para ratio of H2D+, from APEX and SOFIA observations. To date, only one measurement exists, and ours would be the first in the high-mass regime. This makes it possible to confirm the recent result of timescales decreasing with mass, derived from the analysis of the complete ATLASGAL sample.