In 2021, machine learning and AI-driven personalisation are no longer a mere ‘nice to have’, but are the bare minimum of what customers expect from brands. In fact, as Amazon Web Service’s Worldwide Head of Business Development for Applied Artificial Intelligence Zoe Hillenmeyer shared during a recent webinar on the topic, 63% of customers see personalisation as a standard level of service.
“This means when I show up, it had better be recommendation or personalisation, you had better be understanding me when I arrive,” she told the audience during the virtual event. “That’s a really interesting table stake that has become the norm very, very quickly.”
As Hillenmeyer explained, high customer expectations around personalisation have led some marketers to question whether they have enough knowledge around data and AI to truly meet their customer’s demands. Much of this uncertainty is driven by the belief that only those with a deep knowledge of data science can implement AI-driven personalisation.
“There tends to be a feeling that you must have a lot of depth in data science to be able to participate in crafting that experience,” said Hillenmeyer. “That’s the wrong way to think about what’s possible with machine learning capabilities and personalisation. Technology is becoming a bridge between data science and design.”
The best implementations of machine learning and AI-driven personalisation and recommendation, she said, are being crafted by hybrid teams, made up of both creative and scientific minds. “People are learning from one another in a really active agile way,” she said. “It’s very fast, very interactive.”
AWS Webinar on machine learning for CMOs
She pointed to an example from make-up store Mecca, where the brand was able to implement the capacity to personalise its email marketing within the space of a few weeks, which eventually led to a 65% increase in email click-through rates and a corresponding increase in email revenue.
“The reality is that everything is personal, or at least customers want everything to be personal. So it’s about getting the right products, the right images, the right product, title, default categories, product rankings, outbound messaging; the whole thing.”
John O’Mahony, partner at Deloitte Access Economics, who was also on the panel, reminded the audience that data analytics is not simply the same old marketing with some numbers added. “It’s also changing where marketing fits inside organisations,” he said.
In a report from AWS and Deloitte called Demystifying Data, researchers spoke to 300 ANZ businesses to understand their perspectives on data. They discovered that there were gaps in perception around what kinds of data is important inside organisations.
The researchers discovered that just 35% of businesses identified industry or customer research as organisational data; while just 38% of businesses identified call centre recordings or logs as organisational data.
O’Mahony explained that while customers will produce data that will be relevant for marketers, that data will also be important across the entire customer service journey, and can be utilised beyond marketing alone.
Using call centre logs as an example, O’Mahony explained that while this data will clearly be helpful for helping to improve the outcomes for customers, it will also have benefits for other parts of the organisation including monitoring compliance, research, and supporting lead generation.
“One of the frustrations I’ve seen from CMOs is how sometimes their role can be narrow or compartmentalised,” he said. “It can be difficult to explain the benefits of marketing activity. Data analytics and machine learning offer the opportunity to better track what you’re doing, and to be able to get the investments that you need to change the organisation and the marketing function.”
AWS’ Ben Kidney, who joined Hillenmeyer and O’Mahony on the webinar, explained that data gives brands an opportunity to not just ‘say’, but ‘do’. He shared the example of Aussie food brand Tip Top, which was able to make good on its promise to reduce waste, doing so through the use of data.
“The intent from Tip Top was to reduce waste without limiting the physical availability of their products,” he said. “This resulted in a 30% reduction in overstocking and a 10% reduction in understocking. So there was an environmental benefit, there was a huge cost saving, and they delivered fresher products to their customers. That is a really strong marketing proposition, enabled by data.”
So how can marketers start to encourage the entire business function to support investment in machine learning? According to O’Mahony, in order to get funding and business support for these initiatives, marketers must get better at business advocacy.
“A lot of what’s happening will require interaction with other parts of the business that are holding the data, or with finance, in order to get what’s needed for investment,” he said. “In marketing, the first thing we need to be able to do here is to put together the business case, to be given the permission internally in your business to be able to take the investment steps, get your hands on the data, and to execute something that’s small, something that’s doable.”
Finally, for those who remain uncertain of their own knowledge base, Hillenmeyer suggests building a culture where it’s ‘okay to not be an expert yet.’ “My team have learning days once a quarter where we take the day to talk about what we’re working on learning,” she said. “We share, and we encourage each other on that journey.”
Resources such as AWS’ free ebook, Unlock the Potential of Machine Learning for Executives in Australia & New Zealand, are also a great resource for those just starting out on their learning journey.
To discover more insights from the webinar, and to rewatch in full, click here.