Surface-enhanced Raman spectroscopy (SERS) is a highly sensitive analytical technique with excellent molecular specificity. However, separate pristine nanoparticles produce relatively weak Raman signals. It is necessary to focus on increasing the "hot-spot" density generated at the nanogaps between the adjacent nanoparticles (second-generation SERS hotspot), thus significantly boosting the Raman signal by creating an electromagnetic field. This study employed a self-assembly method without using modifiers based on promoter-induced self-assembly to synthesize stable and plasmonically active surfaces from citrate-reduced Ag and Au nanoparticles. Hierarchical structures like Pickering emulsions (PEs) and stable plasmonic aggregates (SPAs) were studied, focusing on controlling their sizes using "promoters" (TBANO3). The sizes of the SPAs were also adjusted from 85.5 nm to 136 nm by regulating the ratio of the water to the oil phase. Furthermore, to understand the distribution of "hot-spots" on these Au or Ag hierarchical structures, the electric field was simulated using the finite difference time domain (FDTD) software. Third-generation hotspots were also created using hybrid structures of plasmonic nanomaterials and surfaces to significantly improve SERS detection by depositing the colloidosome structure on Cu foil (AgSPAs/Cu substrate). The SERS signal was amplified by achieving an enhancement factor of 7 × 107, compared to an enhancement factor of 2 × 106 when using the AgSPA/glass substrate. Significantly, the limits of detection (LOD) and quantification (LOQ) for the colloidosome substrate to detect crystal violet were found to be 4.51 ppb and 13.66 ppb, respectively. The reproducibility of the prepared substrates was demonstrated to be commendably high, characterized by relative standard deviations (RSDs) of 8.00% for the 1177 cm-1 peak, 7.61% for the 1588 cm-1 peak, and 9.35% for the 1619 cm-1 peak. The AgSPAs/Cu substrate's demonstrated reliability made it suitable for detecting and quantifying analytes, potentially for determining trace amounts of pesticide residues. The LOD and LOQ for thiram detection were calculated to be 0.1 ppm and 0.3 ppm, respectively. These findings highlight the effectiveness of increasing electromagnetic field density for SERS enhancement.