However, Unilever is now working with Google Cloud, utilising a wide array of its analytics and AI tools, combined with extensive consumer and social media data, to precisely target its consumers when running campaigns.
This week I got the chance to speak with Alex Owens, VP Global People Data Centres at Unilever, at Google Cloud’s Next conference in San Francisco, where he explained how the company is becoming more agile in its approach and is using Google’s technology to adjust campaigns on the fly to better reflect the needs of consumers in different regions.
Owens described how Unilever has implemented Global People Data Centres (PDC) in 30 markets across the world, which each focus on three core elements. Firstly, social and business analytics, which is a team that generates insights from consumer data and social media. Secondly, there are consumer engagement centres in each PDC. These are call centres, which not only collect data on complaints, but also suggestions from consumers. And finally, the third element is people relationship marketing (or CRM), which is accountable for data driven marketing across the organisation.
Owens said that these three pillars are intentionally together because they deliver “insight through to execution”. He added:
“We have a very lofty ambition of creating 1 billion one-to-one relationships, through meaningful and relevant dialogue. So what does that truly mean? We want to create 1 billion connections with our consumers. And when I say connections, I don’t just mean chucking a bit of content at them. I mean actually building a dialogue and a meaningful relationship.
“We have been on a journey of using consumer data to market better and to make it better for them, over the last two years. We have 2.5 billion people using our products every day. It’s a huge opportunity to capture data. We need to make sure we use that data in an appropriate way, to ensure we retain the trust of our customers.”
Working with Google Cloud
Google Cloud has been a strategic partner of Unilever for many years, initially focused on the ad sales side, given how much the company invests in advertising. However, over the past two years it has been working with Google Cloud more closely, to adopt a multi-cloud approach. Unilever also makes use of AWS and Azure, as Owens said that it likes to have “multiple providers”.
However, the PDCs are making use of Google Cloud’s API offers to better make data driven marketing decisions. So, for example, it uses Google’s translation tool, as a lot of the insight Unilever generates needs to be translated, given the reach of its operations across multiple countries.
It is also using Google Cloud’s Vision API, which allows Unilever to decode all the static imagery it receives from places like Instagram. It is also using natural language processing, because, as Owens explains, Unilever “wants understanding of the multiple insights that we get from multiple platforms”.
He goes on to say:
“More importantly, we’ve been using Google Cloud a lot within data driven marketing. So, how can we use the Vision API to help us understand user generated content that’s been generated as part of a campaign? That allows us to change content if required in order to optimise in real time.
“Because of how much we spend on advertising, Twitter gives us a whole feed of data. Within Instagram we get an API feed that goes into Brandwatch, which is the query tool we use to take all the data and query the data from all the social platforms. And then we use Vision API to decode the imagery and to help us understand all the insights we get from them.”
Unilever also has developed its own internal tools using Google’s Vision API, which allows the company to not only decode imagery, but also to help it understand the situation in which the imagery exits.
What does all this enable?
So, what does all this look like? Owens said that one of the biggest campaigns it launched was around its Close-up toothpaste brand in Asia. Using search analytics, Unilever discovered that the second most searched for term on their website was users searching for the term “learning how to kiss”. Using this, the team created a three-day campaign for the brand around Valentine’s Day, deployed in six key Asian markets.
However, Asia is a diverse region and there are cultural sensitivities to consider when delivering a campaign focused on kissing, romance and relationships.
Owens said that Unilever didn’t get this entirely right from the start, but was able to make quick adjustments in an agile fashion to better target its consumers. He explained:
“We used Google Cloud effectively to take the data consumers were sharing to make sure we were hitting the spot on the campaign, given the cultural diversity of the markets we were operating in. We had to adjust quite a lot.
“Even within Indonesia there is massive cultural diversity and we didn’t quite hit the mark initially. So what we realised is, the way that consumers were talking about it and the way they were sharing user generated content, was that from a diversity point of view, we didn’t quite hit the mark. So we refined and re-tuned the campaign, we did about 200 re-edits to stuff.”
“In the past we would have broadcast to all. Now we can broadcast to segments, or even one on one at an individual level.”
Success and challenges
Owens explained that Unilever is measuring success in a number of ways. Firstly, is Unilever making those one billion connections it set out to achieve? It is also is considering whether data driven marketing provides a higher ROI than traditional media, such as TV.
However, Owens was happy to admit that Unilever wasn’t the first to the game to get this right and is still undergoing the changes needed to execute this successfully in the future. He said:
“We are an organisation of complexity. But never underestimate Unilever’s skill. We are now understanding what we need to do and how we need to operate differently. We know we need to make changes if we are going to survive in the future.
“And therefore, I think, if you look at that over the past three years since I’ve been at Unilever, we were in five markets doing data driven marketing at scale, we are now going to be in 24 by the end of this year. Within that, agility becomes important.
“On the tech side we have to move away from waterfall to agile. You need sponsorship from the top in order to execute, but equally we need the right skills in the organisation to operate this at scale. So we will be bringing new people in, as well as upskilling current people.”
However, this doesn’t come without its challenges. Namely convincing people that this change, these tools and the new data driven approach is effective. Also, skills are always a concern. He said:
“You’ve got to simplify in a way that people understand. It’s complex and many people don’t truly know what it looks like. And therefore until they truly feel it and see it, they don’t understand what it should be.
“The other challenge is keeping up with the pace of the change. I think now that the organisation has got ‘it’, everybody wants it at the same time. Also, the tech landscape changes so frequently and so much. The only constant is change.
“And finally, bringing new skills into the organisation that are quite scarce, a scarce commodity, is front of mind. They’re either in high demand in the US, or are limited in other markets.”