Advertisement

News and analysis on the global poultry
and animal feed industries.
on July 1, 2009

Benchmarking becomes a numbers game

Nobody should be surprised that a survey of breeding results in sow herds this year is reported to have struggled with the poor quality of performance records on many of the units.

Good recording in a herd of commercial size takes time and a close attention to detail, even with modern electronic assistance, so the job is often poorly performed. Errors are common.

Let me confess that I put myself among those who are not good at keeping accurate records. So, it is no criticism from me to say that the figures generated for a typical unit can be expected to be inaccurate. But nor is it an excuse, when the fact is that we now need our data to be correct so we can manage through the melee in higher feed prices and other costs.

There is an equal problem waiting for anyone who wants to check the figures of competing firms as a way of benchmarking their own performance. How do you know their basis for recording is the same as yours, so the results are comparable?

Definitions differ

Another reminder of this has appeared recently in the observation that the countries belonging to the European Union do not always agree on when a female breeding pig should be included in sow numbers. The composite data reported for the EU-27 considers maiden gilts to be sows for counting purposes, whereas national systems may begin only with gilts after mating. In some cases it has been making a difference of almost 10% in the apparent sow inventory, with obvious consequences for any calculation of productivity purporting to show pigs per sow per year.

On a narrower level, the rise of artificial insemination for pig breeding in recent years has brought a further possibility for confusion. Herds that still maintain their own boars may well continue to include the feed given to those animals in the total fed per sow per year, for example, whereas the boar contribution is absent in the case of AI herds. Maybe they should add a notional amount so that their feed efficiency can be compared more accurately.

On a similar matter, it was interesting to see a set of records in one country this year that included a provision for AI charges. It worked out at only 1% of the total production cost per pig and so half of that needed for, say, handling manure, but all the same it is not an entry found yet in many other herd cost listings.

Labour grows in significance

In terms of comparability, labour charges have become really contentious. Years ago labour represented a rather trivial proportion of total costs per pig. Now, it is an important matter, with the amount paid to managers and workers possibly representing 15-16% of all costs. This seems set to rise again as wage rates increase to take account of the higher living costs. Food is costlier than a year ago, as is oil. Inevitably, there will be firms trying to economise by having fewer workers for the same pig numbers. A safe prediction is that they will fail, if they want to do this while still maintaining productivity.

How we measure the labour input is itself debatable. Most of us tend to work on the long-established value of sows per worker. Modern advice is that we should change this to a calculation of the number of man-hours spent per sow in the herd. We are told under 15 man-hours per sow is good for piglet production, with 25 hours (probably more) to take the pigs through the nursery and grow-finish stages. However, all depends on the housing, the herd size and the manure system. In my view the sows per man standard is much easier to calculate and to compare.

Feed prices have greatest impact

All this looks miniscule when set alongside the big numbers that are for feed. Say it accounts for 65% of your per-pig costs. It must be the first item to consider in any efficiency drive. When we look around for guidance on what is possible we hear, for example, that the Dutch and the Danes can achieve a liveweight feed conversion of under 2.70 where most of the rest of us have to accept no better than 2.80 or even 3.0. This ignores the question of whether to calculate feed efficiency ratings only from weaning or from grower entry, or to go for a whole-herd evaluation.

Calculating a herd feed conversion rate (FCR) is easyyou just take the quantity of feed purchased or made for all the pigs on the farm and divide it by the weight of pigs sold. The result will vary according to the sale weight of the animals, but good managers are achieving around 3.5 to 110 kilograms or close to 3.2 for pigs sold at 100kg. Personally, I am less comfortable with the idea of figuring out the amount of sow and piglet feed used for every pig reared. Advocates of the system reckon efficient use means less than 9kg feed for every kilo of piglet leaving the site. I have yet to look at the number and recognise immediately if it is good or bad.

My difficulties in interpretation could increase greatly, if we go along the lines that some people propose of taking energy conversion rate (ECR) rather than FCR. They say energy is the all-important element in cost control and we should set a target of about 44 megajoules for a herd ECR measured in digestible energy. They may well be right. Just do not expect me to agree or disagree until I have had time to work through some numbers. Benchmarking cannot be this hard, can it?

Comments powered by Disqus