Different low-cost cobalt precursors (acetate, chloride) and thermal treatments (air calcination/H2 reduction versus direct H2-activation) were investigated to alter the interaction between cobalt and silica. H2-activated catalysts prepared from cobalt chloride had large Co0 particles (XRD, chemisorption) formed by weak interactions between cobalt chloride and silica (temperature programmed reduction (TPR), TPR with mass spectrometry (TPR-MS), TPR with extended X-ray absorption fine structure (EXAFS) and X-ray absorption near edge spectroscopy (XANES) techniques) and retained Cl-blocked active sites, resulting in poor activity. In contrast, unpromoted Co/SiO2 catalysts derived from cobalt acetate had strong interactions between Co species and silica (TPR/TPR-MS, TPR-EXAFS/XANES); adding Pt increased the extent of the Co reduction. For these Pt-promoted catalysts, the reduction of uncalcined catalysts was faster, resulting in larger Co0 clusters (19.5 nm) in comparison with the air-calcined/H2-activated catalyst (7.8 nm). Both catalysts had CO conversions 25% higher than that of the Pt-promoted catalyst prepared in the traditional manner (air calcination/H2 reduction using cobalt nitrate) and three times higher than that of the traditional unpromoted Co/silica catalyst. The retention of residual cobalt carbide (observed in XANES) from cobalt acetate decomposition impacted performance, resulting in a higher C1–C4 selectivity (32.2% for air-calcined and 38.7% for uncalcined) than that of traditional catalysts (17.5–18.6%). The residual carbide also lowered the α-value and olefin/paraffin ratio. Future work will focus on improving selectivity through oxidation–reduction cycles.