How do you democratise consumer insights in pursuit of ad effectiveness? For PepsiCo, the challenge was not just getting access to those insights but sharing them across a huge, global organisation to benefit future marketing work.
Stephan Gans, Chief Consumer Insights and Analytics Officer at the global food and beverages brand, explained that while the organisation had access to a wealth of consumer insights, “[It] felt like we were absolutely not leveraging our scale, and as a result of that, we were not driving any competitive advantage for our company.”
At the Festival of Marketing (October 3), Gans and Kate Schardt, VP, Global Insights Capabilities & Partnerships, discussed how they went about implementing a new insights and ad testing platform – and how the business tackled the cultural hurdles of bringing people on board with the new technology.
PepsiCo worked with Zappi to create a single platform and a standard and digitalised toolkit that every insights specialist and marketer within the organisation could use – a “connected ecosystem” of consumer insight. The new insights platform was named Ada, after computing pioneer Ada Lovelace.
The new technology could bring together data from any ad tested within PepsiCo and make it available across the organisation; Gans explained, “Every piece of data gets fed into the same database … So, if Gatorade develops 23 different pieces of advertising in a year across the world, we can mine all that data … to get smarter. So that the 24th ad is even more effective than the 21st one.”
To begin with, even the act of working with insights in a digital platform was a major change for the business. Schardt recalled, “We had broad alignment across marketing and insights and in the wider organisation that the way we were working previously was not good enough … Everyone, rationally, was on the same page: We have to get better.
“But there was so much resistance to change because [there are] a lot of stakes in marketing around things like creative testing. … And we were fundamentally challenging people to level up; to get better.”
Zappi CEO and Co-Founder Steve Phillips agreed that, “The technology [of the new insights platform] wasn’t the hard part – it was much more about process, and very specifically, people – and the behaviour change associated with that.”
However, a learning dimension to the new tool aided its uptake: Schardt explained that, “The new tool had a lot more diagnostic learning, so it helped people get smarter and feel like they could be more effective with their marketing.”
When stewarding a new piece of technology and way of working through a business, Schardt emphasised the need to push forward, but with empathy – making sure people feel that their concerns are being heard. “My advice is … you have to work the problem; you have to work the change management programme.
“…You have to meet with the right people, hear what they’re saying, have empathy for why this is a struggle for them … And eventually, when you can convince people, ‘This is going to make me better at my job’ – when they get that first win in the new system – that’s when you’re through.”
PepsiCo and Zappi are now looking to generative AI as a tool for gaining quicker access to insights from within the connected platform – but also generating new concepts for advertising at a speed that wouldn’t have previously been possible.
As Gans explained, “If you want to advertise a brand like Walkers effectively in the UK market, for example… you need to create hundreds, if not thousands, of iterations [of an ad], depending on how sophisticated you want to be – the company doesn’t have enough people to do all that work.
“So you need to find ways to automate, and you can only automate smartly if you have the data at your fingertips.”
Phillips added, “When I started out in consumer insights, our job was to help optimise and validate ideas – ‘Is this going to be a good ad, a good brand, or a good product? Tell me whether it’s good, and also give me a little bit of help on whether I could do it better.’
“When you have a system like Pepsi’s, you can ask different questions – instead of, ‘Will this ad work with young men in Thailand?’ we can ask, ‘What type of advertising works with young men in Thailand?’ And so inspire marketers.”
Zappi has begun using generative AI on top of clients’ consumer insights to generate advertising concepts at speed – “which require a lot of human refinement,” Phillips said, “but you’re creating ideas in minutes or hours and not weeks or months.”
With that said, a lot of underlying work has gone into building up PepsiCo’s data ecosystem to the point where these types of insights are possible – as evidenced by the change management process described by Schardt.
Speaking to that journey, Schardt said, “You have to start somewhere … Pick somewhere to focus – so for us, that was ad testing – and be relentless against the goal. There’s going to be inevitable setbacks; [but] we were all committed to making this work.”