Metal nanoparticles exhibit physical and chemical properties that differ significantly from macroscopic metal phases. In particular, silver nanoparticles (NPs) had have extensive attention since it has a strong interaction with light due to their high quality localized surface plasmon resonances, which can be tuned all across the visible to the NIR regime. These strong and tuneable interaction of visible light with silver NPs have found many applications, in sensing, surface enhanced Raman spectroscopy, photo(electro)catalysis, and more.Polycrystalline indium tin oxide (ITO) films are highly desired substrates for many of these optoelectronic applications. In many of these applications, it is essential to reliably and precisely control the size and density of silver NPs. Even though it is well known that the electrochemical control over nucleation and growth of silver NPs on foreign substrates highly depends on the surface properties of the substrate, yet, the only parameter that is often specified for these ITO substrates is a bulk property: sheet resistance. As a result, nucleation and growth on ITO is highly irreproducible.In this work, we show that ITO substrates with same technical specifications (i.e. sheet resistance, light transmittance and roughness) and supplier may still have different surface properties such as the crystalline texture. We find that these differences in the surface properties has a strong impact on the nucleation and growth. Even though we used seemingly equal ITO substrates, XRD and EBSD scans clearly differentiate between two types of films based on the ratio between the (222) and (400) related peaks. We controlled the density and size of the silver NPs by making use of the double pulse technique. In this method, where the high overpotential nucleation pulse controls the nucleation density. Previous studies have reported an empirical exponential relationship between island density and nucleation pulse overpotential, where the slope is a measure for the electron kinetics.We find that the preferential presence of lower index surfaces (e.g. (100)) leads to few orders of magnitude lower particle density compared to the preferential presence of higher index surfaces (e.g. (111)). Furthermore, the density on the lower index surface is strongly dependent on the nucleation pulse overpotential, while the higher index surface is barely affected. We find that this difference in the relationship is caused by mass transport limited nucleation. We have verified this through a spatial correlation analysis using the nearest-neighbour distance. For higher index surfaces, the nucleation density is already large compared to the Nernst diffusion length at low nucleation pulse overpotentials, such that the limiting mechanism becomes mass transport instead of electron kinetics. We therefore elucidate that the empirical exponential relationship between island density and nucleation pulse overpotential is not enough to draw conclusions on electrode kinetics, where additional validation of the growth process is essential.Beyond careful control of electrochemical parameters, here we show that neglecting the macroscopic surface characteristics of polycrystalline substrates can lead to major ambiguity in nucleation and growth dynamics of metal nanoparticles. We show here that XRD characterization prior to deposition is a simple additional measure to tailor metal nanoparticle electrochemical growth on ITO.