This report draws on 40 months of continuous monthly measurement of LinkedIn posting behaviour, from March 2023 to June 2026. Each monthly snapshot recorded, for each company: total LinkedIn follower count; total employees on the platform; number of employees who posted content in the preceding 30 days; and equivalent figures for CXO-level leaders, VP and Director-level leaders, Business Development and Sales professionals, and geographic subgroups.
860 companies were included in the analysis. The selection criterion was that each company appeared in data from at least two distinct calendar years, ensuring sufficient longitudinal depth for trend analysis rather than snapshots. Companies appearing in fewer than two calendar years were excluded from all correlation and trend calculations.
For each company-month, posting rates were calculated as: employees who posted ÷ total employees × 100. The same formula was applied for CXO, VP/Director, and BD/Sales groups. Rows were excluded where the calculated rate exceeded 100% (indicating a data anomaly), was negative, or where the denominator was zero. After cleaning, the dataset contained approximately 27,000 valid company-month observations across the four primary metrics.
All correlations are Pearson r values, calculated using Python’s scipy.stats.pearsonr function. For each company with six or more valid observations for a given metric pair, the two monthly time series were extracted and Pearson r was calculated. Companies where either series had zero standard deviation were excluded.
the figures quoted throughout this report (e.g. VP/Director → Employee r = 0.676) are the median of all individual company r values - not a single correlation calculated across all pooled data. This is standard practice for panel data analysis and prevents larger companies or those with more months of data from disproportionately influencing the aggregate figure.
For each company: the company’s own median CXO posting rate is calculated; months are split into above-median and at-or-below-median CXO months; in each group, the proportion of months where employee activity exceeds the employee median is calculated; the multiplier equals the above-median rate divided by the below-median rate. The 1.71× figure is the median across all qualifying companies. It holds in 82.5% of companies (i.e. is greater than 1.0 in 82.5% of cases).
The 4× figure compares employee posting activity at companies in the top quartile of VP/Director activity with those in the bottom quartile, across all pooled company-month observations. This is a cross-sectional comparison and partly reflects differences between high-activity and low-activity companies overall, as well as within-company variation. Both the multiplier and quartile ratio are reported as they capture different aspects of the relationship.
Each company was manually assigned to one of 11 industry groups based on primary business activity. Industry-level correlations use the same within-company methodology, restricted to companies within that industry, and are reported as medians. Leaderboards and industry analysis exclude sub-brands, regional entities, companies with fewer than 1,000 LinkedIn employees, and companies with fewer than 12 months of data.
All analysis was conducted in Python 3 using: pandas (data loading, cleaning, panel construction), numpy (array operations), and scipy.stats (Pearson r, p-values, linear regression). No sampling was applied; all calculations used the full 860-company panel.
Full methodology and the underlying dataset are available on request. Contact us at hello@tribalimpact.com