Quantifying the value of on-farm measurements to inform the selection of key performance indicators for livestock production systems

Abstract

The use of key performance indicators (KPIs) to assist on-farm decision making has long been seen as a promising strategy to improve operational efficiency of agriculture. The potential benefit of KPIs, however, is heavily dependent on the economic relevance of the metrics used, and an overabundance of ambiguously defined KPIs in the livestock industry has disincentivised many farmers to collect information beyond a minimum requirement. Using high-resolution sheep production data from the North Wyke Farm Platform, a system-scale grazing trial in southwest United Kingdom, this paper proposes a novel framework to quantify the information values of industry recommended KPIs, with the ultimate aim of compiling a list of variables to measure and not to measure. The results demonstrated a substantial financial benefit associated with a careful selection of metrics, with top-ranked variables exhibiting up to 3.5 times the information value of those randomly chosen. When individual metrics were used in isolation, ewe weight at lambing had the greatest ability to predict the subsequent lamb value at slaughter, surpassing all mid-season measures representing the lamb’s own performance. When information from multiple metrics was combined to inform on-farm decisions, the peak benefit was observed under four metrics, with inclusion of variables beyond this point shown to be detrimental to farm profitability regardless of the combination selected. The framework developed herein is readily extendable to other livestock species, and with minimal modifications to arable and mixed agriculture as well.

Towards regenerative agriculture: the case of dairy and arable farming in the Netherlands

“…While widely applied in the health and education sectors, their use in agriculture is currently in a pioneering phase while the suitability of such instruments for the sector is debated (Janus and Holzapfel, 2017). A common pitfall is the application of result-and outcome-based metrics in isolation from practices, this hampers the interpretation and understanding of monitored results and outcomes when the context is not identified (Jones et al, 2021;Wade et al, 2022). For example, when only soil carbon is monitored to track progress towards climate goals, it remains unknown whether this progress can be ascribed to the original intervention or to other factors (e.g.…”