This study presents a novel way to measure the contribution of private extension to farm productivity for club data. Club data refers to any convenience sample obtained from a study group, consulting firm, cooperative or producer organisation. The study develops a stochastic frontier production function model with the inefficiency effects of pasture-based dairy farming in the Eastern Cape, South Africa. The analysis for 2012–2018 involves 49 adopter farms, and controls for inter-calf period, nutrient use efficiency and the amount of extension contact. Results are robust to functional form specification and there is no evidence of frontier-shifting technical progress for the Cobb Douglas or translog model, but there is a clear productivity benefit to engaging with the private extension service provider working locally (adoption). Productivity rises at 0.91–1.06% p.a. over time and by 1.54–1.62% p.a. with each extra year of the extension. Large farms close to the private extension provider’s base of operations benefit most from being in the group. This case study is important because it documents productivity growth in the period since 2010 and puts the effect of extension on productivity growth back on the local research agenda.