To Benchmark or Not?

Use of Livestock EID for Benchmarking Analysis

This is number three in a series of articles about practical application of EID (electronic, or radio-frequency identification of animals) in pastoral livestock farming. If you haven’t already, you may wish to read the first and second articles.

If you’re farming, you may already have applied EID tags to your animals to meet your country or region’s traceability requirements. You may have even purchased tag reading equipment and recording software in order to be part of a food chain information or farm assurance programme. If that’s the case, you’re no doubt wondering if you can leverage this technology to deliver continuous improvement to farm profitability.

One of the pillars to continuous improvement is benchmarking. In farm systems meaningful benchmarking compares similar resources with differing management and technologies. If done well such analysis can highlight which management and technologies are achieving a superior result. You will often gain substantial benefit in this way by learning your strengths, weaknesses, and opportunities for change.

EID can be a powerful tool for tracking the performance of individual animals where individuals in a mob have different treatments or genetics. For example:

  • You want to know how progeny from your regular ram supplier performed against the new breeder you used this season;
  • You finish cattle from a number of farms and want to know who you should buy more from next season (and who you should avoid); or
  • You may be looking at the performance of animals who you treated using a different drenching regime, or that you fed for a time on different forage.

Recording animals tagged with EID tags can be a powerful tool to let you sort and filter the data to get answers to all these questions. If you capture data at many points along the way, you have a substantial data set with a variety of variations you can analyse!

But there’s the rub as well. It is easy to analyse data from animals that have been treated in a number of different ways, and end up not understanding which treatment caused the difference. Did the animals perform poorly because of the new forage, or because I didn’t calculate the break size and underfed them? Did the new ram team perform well, or did I mate them to heavier ewes? When lambing in separate paddocks so I could apply EID tags to the groups, did the post-lambing management impact more than the genetics?

I’m an enthusiast for the role of this sort of learning exercise to identify opportunities and encourage farmers in trying new things, but as my colleagues point out, there are challenges with properly designing your “benchmarking experiments” and capturing the data necessary to interpret the results.

This is what those in the know call “experimental design”, and it is the reason industries pay scientific research companies to plan and oversee research trials. If you’re in a region where research trials have already been run on the forage or breed you’re considering, should you repeat it with your own trial programme, or use what the experts have already uncovered?

Research trials strive to narrow down the variation caused by the treatment versus variation caused by the system. To achieve this they ensure animals receive the same management in all aspects except the treatment being examined. Trials are replicated so that statistical analysis can differentiate variation due to the treatment being studied from the inherent variability in the system. Our on-farm trials tend to be more limited, more “retrospective”, and can have less power for predicting future outcomes, that is are we sure the result can be repeated?

This is not to say that you shouldn’t use EID in benchmarking at all. After all, if you have the equipment and there are valuable things to be learned, it is worth investing some effort in doing the job properly. Here’s a list of some things that I think are worth considering:

  • Identify what you are hoping to learn, and why it will be valuable to you.
    What is the question? Is it whether per head performance will be greater on a particular forage or is it the total meat production per hectare? There’s going to be some effort in planning, managing, collecting information and analysing it. You want to be sure that this is a useful lever to add value to your farm, and not just overhead. If in doubt, start with small, simple questions.
  • Plan your trial design. Identify how you could end up with poor data and plan against it. Think through what results would give a definitive answer. Would it skew your results if you put all the heavy animals onto the new forage, and the light animals onto grass? If so, decide how you will randomise those weights. Will you be able to analyse the data if you don’t weigh animals into and out of groups or paddocks? If you need those weights, plan how you will get them.
  • Keep great records. 
    Think through how you will manage the data. The majority of farmer trials never get analysed because the data was not recorded correctly. Often farmers end up with various files with short 8-character session names or just dates and numbers, and struggle to communicate what was recorded or done in each session. This is one of the key reasons for moving from simple scales indicator and files to fully-fledged recording software – but you can make poor session names in recording software too. Keep a diary or notebook so you can make sense of things if needed.
  • Do the analysis and act on the results.
    You’ve invested time and energy bringing animals into the yards and weighing and sorting them. If you don’t do the analysis, it is a waste. Choose good software that helps you, and if necessary ask for help.

There can be value in benchmarking or analysing how groups of animals perform using EID, but it takes time, effort, and good planning. It’s only worth doing this for things that you can control, and which can have a material future impact on your farm.