Speaking at Leatherhead Food Research’s food innovation day last week, Dr Guillermo Hough said that consumer testing on products at the end of their shelf-life typically revealed that “many could have had their shelf lives extended”.
Hough, a research fellow at the Instituto Superior Experimental de Tecnologa Alimentaria in Argentina, said that many products had a relatively low rejection rate from consumers at the end of their shelf-life, which suggested manufacturers were being overly conservative with best-before dates and good food was going to waste.
Given that many products were typically consumed well before their best-before dates, even a 50% rejection rate at the end of a product’s shelf-life might be acceptable given the small numbers of consumers that would eat the product at this late stage, he added.
Using the example of a product with a 35-day shelf-life consumed by 100 people, only four would typically consume it after 20 days, he said, meaning that a 50% rejection rate at the end of the shelf-life only correlated to two consumers. And this was “probably acceptable”, he said.
Quality, acceptability, not food safety
He added:“With cheese, meat, milk and so on, it’s microbiology that dictates shelf-life, but in 80% of categories, it’s sensory shelf-life that matters, not food safety, so the issue is consumer acceptance.”
However, it was still fairly unusual for food manufacturers to use consumer testing in shelf-life work, he said. Most firms instead relied on a combination of historical information, instrumental/analytical studies and trained sensory panellists to predict shelf-life and set best-before dates.
But correlating consumer rejection rate data with data from trained sensory panellists and instrumental data might help firms select a more appropriate cut-off point at which a product’s shelf-life should be set, he argued.
For example, once a certain flavour compound known to be a marker of rancidity/fat oxidation reached a particular level (as measured by analytical techniques such as GCMS), or once sweetness (as measured by a trained sensory panel) reached a certain threshold, consumer acceptability might suddenly drop, with rejection rates hitting 50% or more.
Conversely, if trials repeatedly showed a correlation between sensory/analytical data and consumer rejection rates, it should be possible to use such data to predict consumer acceptability.
If a firm decided that a 30% consumer rejection rate was the point at which it wanted to set a product’s shelf-life, it could predict, from sensory panel or analytical data, when a 30% rejection point would occur, and set the shelf-life accordingly.
Accelerated shelf-life testing
Hough also sounded a note of caution about the use of accelerated shelf-life testing, a widely used tool to help manufacturers predict shelf-life.
But trying to predict what a product would be like after nine months after exposing it to ‘extreme’ conditions over, say, three months, was inherently risky, said Hough.