Lex Localis (2025) — Social media marketing, engagement, CX, and growth

Summary:

This academic paper, published in Lex Localis – Journal of Local Self-Government by researchers from five Indian universities, tests whether Facebook post choices (post type, paid promotion and posting time) are linked to engagement, deeper content use and reach. The authors propose a simple chain for SMEs and local governments: strategy → engagement → customer experience → growth.

The study analyses 20 Facebook posts. The paper describes the dataset as simulated, modelled after open-access Kaggle data, so it is not real business or government performance data. The authors ran descriptive stats, Pearson correlations and stepwise regressions. Engagement is measured as likes, comments, shares and engaged users. Customer experience is measured as “consumers” and “consumptions”. Growth is measured only as reach and impressions.

Key findings:
• Paid posts reached more people: an average reach of 3,221.692 for paid posts vs 2,449.714 for organic, and 4,358.692 vs 3,628.143 impressions.
• Video posts had the highest average reach (4,192.5) and link posts had the most interactions (375.2). Status posts were lowest on both (2,249.714 reach, 205.2857 interactions), even though the paper’s text calls status the reach leader.
• Engaged users moved closely with the experience measures: r = 0.785259 with consumers and 0.803959 with consumptions.
• Engagement showed almost no link to reach in the correlation table (r = −0.06517 for engaged users vs reach). This sits poorly with the paper’s claim that engagement drives growth.
• Monday posts had the highest engagement rate (0.303951) and Thursday the lowest (0.080899). The paper’s text says weekends, especially Saturday, performed best, but Saturday’s rate was 0.102141.

Margin notes:
Read every finding with care. The data is a simulated set of 20 posts, the design is cross-sectional, and the paper uses causal language (“drive”, “influence”) that this design cannot support. The regression results are never shown, so the partial-mediation claim cannot be checked. Several numbers in the text don’t match the paper’s own tables: reach–impressions r = 0.90 in text vs 0.951105 in Table 5, interactions–consumers 0.72 and 0.76 in text vs 0.57461 and 0.588392 in the table, and a mean engagement rate of ~0.11 in text vs 0.141007 in Table 2. “Growth” here means reach and impressions only, not sales or service uptake.