Up until recently, the main benefit of digital data for advertisers has been a near real-time feedback on campaigns. This has allowed brands to learn from their measurement to continually improve performance.
However, this is changing. Data is also starting to be used to improve the nature of products and services — not only in their campaigns. It’s a case of making the new data actionable and useful rather than just a fascinating insight for the next campaign.
Inevitably China leads in this new trend. Their 500m+ online population can provide so much data for brands in purchase behaviour and comments. For example, Mars recently introduced a new chilli variant of its Snickers bar in China, partly created because of insights gained from looking at thousands of product reviews and comments in forums that suggested that a spicy chocolate bar would be popular in China.
Another example of this is Choosy, a fast-fashion start-up with offices in New York and China. They use Instagram data to help design its fashion collections. They have developed a way to analyse the response to celebrity and influencer pics to work out what trends are rising, and then produce similar items. It’s not just a case of looking to see what gets the most likes. Their algorithms also factor in real potential demand by searching for phrases in the comments like ‘Where can I buy this?’ to separate people who like and want the look from people who merely like the personality.
Other Chinese examples that we’ve seen include the creation of new cosmetics and consumer electronics based on learnings from thousands of online reviews and comments.
Despite China leading the trend, this trend has also been seen at Spotify. According to Daniel Ek, Spotify founder, rock band Metallica uses its streaming data to see the most popular songs in different towns where they play and adjust their set list to make sure that they play their best set. While we don’t know the full details, it’s unlikely that they would play an entirely different set for each city, but instead, it seems very smart to use this sort of data to make sure that they pick a couple of songs from their back catalogue each night to delight local fans. In this way, they are using data that did not exist ten years ago to optimise what they are presenting to fans.
This is a new trend and a unique opportunity — it’s a practice currently at the margins so far — just a few people doing it — but it is an exciting extension of what is possible.