Firms should go beyond simply collecting large amounts of information, and look at how data can be used to make important business decisions – instead of acting on “gut feel”, said Rupert Naylor, senior vice-president for Europe at Applied Predictive Technologies.
The solution to making better use of data, he said, lay in “in-market business experimentation”.
“Rather than ‘how many [page] impressions did we receive?’, companies should be asking ‘what was the sales impact, by category, of our new ad campaign?’,” said Naylor.
“The key rests in implementing each new programme, such as a product introduction, promotion, or marketing campaign, in a subset of retail partner stores or markets.”
Cannibalisation of existing products
Through this method, producers can move from ‘what are the sales of our new pack size?’ to ‘did we see a net increase in sales, inclusive of cannibalisation of existing products, and in what types of stores did it work best?’, he suggested.
Testing each business idea in this way allows firms to isolate their net impact, despite external factors and the “noise inherent” within big data, Naylor added.
When conducting in-market experiments, he said each test should have a well-defined business objective, and that the test results should be segmented so that each programme could be tailored and strategically targeted for maximum impact.
“In-market testing enables food manufacturers to easily and efficiently distinguish between winning ideas and losing ones,” Naylor explained.
“In contrast, companies that are content with only an overall test result are not fully leveraging their big data and run the risk of missing key opportunities.”
The unfulfilled potential of big data has also been highlighted by software provider InfinityQS.
Data trapped in local databases was preventing organisations from taking a strategic view of their operations, claimed the firm’s Martyn Gill, md for Europe, the Middle East and Africa.
“Many manufacturers not only use siloed data, but also have cultures that deal with problems exclusively at a local level, focusing on individual plant performance.
“This prevents them from analysing data on an enterprise-level and, therefore, limits their ability to gain real-time visibility over larger systemic problems.
“Deploying new data collection practices and modern analytics upon a backbone of strong quality software will allow organisations to gain visibility over disparate data and leverage them to drive strategic thinking across manufacturing operations.”