Terra Technnology's 2012 benchmarking study on forecasting showed a 1-percent increase in the forecast error rate among the large CPG companies surveyed, which followed a 2 percent increase in 2011. "We clearly are not headed in a good direction," says Byrne.
The annual Terra Technology survey measures trends among users of Terra's forecasting software, which include 13 of the world's largest CPG companies. Byrne notes that these companies represent, on an annual basis, about $100bn in sales, roughly four billion cases, 90,000 stocked items and almost 500 locations. "It gives a pretty comprehensive look at what is happening in consumer packaged goods," Byrne says.
Other highlights of the 2012 survey cited by Byrne are: generally flat overall sales, a considerable increase in promotions, and an increase in net sales. "It looks like these companies are having to run a little just to stay in place," says Byrne. "I think they also are trying to recapture some of the share lost to private label and store brands during the recession."
One of the key things companies can do to turn this around and decrease forecast error is to make better use of data from retailers to refine their forecasts, Byrne says. He compares the situation to a trip to the store for milk. "When I need milk, I don't think about how much milk I bought this time last year - I look in the refrigerator," he says. "Looking in the fridge is far more effective." Similarly, analyzing point-of-sale retail data is the most effective way to reduce forecast error, he says, noting that Terra's customers who take this approach experience, on average, a 30-percent to 50-percent improvement.
Retail data has been available for a while now, but the volumes are daunting. "If you just think about Walmart, it has 3,000-plus stores and thousands of active items, so you are looking at millions of transactions a day," Byrne says. "It is up to the manufacturer to take that data and use it to improve the supply chain."
Byrne says that Terra's demand-sensing technology provides a structured, robust way to do that by applying the right mathematics. "It is really a problem of scale," he says. "You need the computing power."
Looking to the future, Byrne sees continuing pressures on costs and cash flow for CPG manufacturers. One positive aspect is that this drives innovation, he says, noting that the 2012 survey showed a significant increase in new products. "I think we will see more innovation, which will drive more volatility and more supply chain challenges."
Byrne also predicts "a re-imagining of VMI over the next year or so based on more granular, more accurate data." To support this "there will be a big push for better, more automated out-of-the-box solutions from vendors," he says, adding that manufacturers should hold software companies "a little more responsible for improving the quality of their output."
