This study presents a novel spatio-temporal rental index designed to analyze rental trends in the commercial office markets of major Indian cities. Moving beyond conventional price-based indices, which often suffer from biases caused by simple averages and outliers, our approach employs a Bayesian spatio-temporal modeling framework to capture nuanced market dynamics. By incorporating both actual transaction data and predicted hypothetical transactions for active leases, the methodology overcomes data sparsity challenges, offering a comprehensive and accurate representation of market trends. A key innovation of this methodology lies in its efficient handling of new data. The model updates the index quarterly using data from the current and the previous three quarters while leveraging previously fitted posteriors as priors. This approach significantly reduces computational overhead without compromising accuracy, as validation against full-data models consistently confirms robust results. The resulting index provides actionable insights at various levels of granularity, including city, macromarket, and micro-market scales, making it a valuable tool for investors, developers, tenants, policymakers, and real estate investment trusts. By reflecting true market patterns and enabling informed, data-driven decisions, this index sets a new standard for analyzing commercial office rental dynamics and advancing market intelligence.