Renault is letting machine learning power the decisions it makes about influencer marketing.
The automaker is testing an AI social listening tool in a number of European markets which allows it to easily segment audiences and identify influencers they trust.
It's part of a wider shift by the brand to move away from mass reach and towards one-to-one communication on social, a strategy being led by Francois-Xavier Pierrel, corporate director of data, CRM and social.
Pierrel was poached from Facebook last year to sit in the newly-created role, and admitted it was “scary” that the car marque didn’t have anyone in this position with a firm grip on its data until just last year.
However, now that he’s settled in, it’s clear he’s trying to break down internal data silos and realign the company’s social strategy to both engage and retain customers.
Using smaller, but trusted, influencers
As part of this, the brand has combined eight of its internal databases – which comprise everything from dealerships to connectivity and battery usage – with a fresh product from Socialbakers that allows it access more detailed and actionable insights: including the social content Renault’s online audience responds to and the influencers they like.
Renault has been using the beta version of the machine learning-powered tool for a few months now in order to help it single out smaller influencers with around 3,000 followers. It can also use the tool to adapt its own tone of voice when it's trying to appeal to more niche audiences.
It's still early days, and while he didn't share specific results Pierrel said he has been “stunned” at how quickly the move has delivered value for the brand.
"Traditionally, brands like us go to David Guetta or people who are very visible... but this is helping us more to influencers with less followers but who are of better quality," he told The Drum during the Socialbakers Engage conference in Prague last week.
Pierrel conceded that the effectiveness of these campaigns is chiefly measured through 'engagement' metrics (ie. likes, comments) at the moment. However, he said the brand was striving towards was a system where it could measure how social drives conversation rates, as well as how it was impacting the way audiences interact with other mediums, like TV.
"This is a much wider [model of] attribution... to make sure that we understand where the money is going, and how much is generated as a kickback," he noted.
The marketer asserted that typically whenever Renault takes a new product to market it relies on its creative and media agencies to come up with this kind of data, which “costs quite a lot of money”.
Renault’s new way of working takes a lot less time, is easier to scale and potentially does away with the need for third party agencies. But in a world where the constant chatter of in-sourcing and efficiency cuts has led to agency jitters, Pierrel was explicit that this wasn’t part of some plot to completely write agencies out of the equation.
While he did admit the brand has been scaling up its in-house capabilities on the data front, he stressed that Renualt was handing over these social insights to agencies, rather than actioning it themselves.
Bringing 'the brain' in-house
Pierrel acknowledged that bringing more data expertise in-house has helped address any knowledge gaps in the business, making Renault more likely to call out discrepancies on the digital and media side.
Media agencies he added, give Renualt "stronger legs" in the many markets is operates in.
“We will never have an internal agency, well maybe one day but for now it’s not the case," he mused, "but what is good, is instead of having all the brain outside, we have the brain inside and we have the natural inclination to internalise some stuff so we know where our money is performing better."