“Consumer expectations have sky-rocketed to the point that hyper personalization is no longer optional for brands, it’s imperative,” said Adobe’s Aseem Chandra this week. This is a clear trend we have been observing for some time. Consumers demand tailored experiences to their individual preferences. And this is dividing the consumer-facing business landscape into two groups: those who can deliver a personalized experience; and those who cannot.
When a CEO is faced with the conundrum of how to deliver personalized service, the starting point is always data. What do you know about your customer? About their preferences? Their needs? Their spending power? Their creditworthiness? And not just today; going back in time. And what about tomorrow? Do you have channels to continually gather customer data as it evolves with changes in your customer preferences?
Without data, the ability for a business to tailor offerings to the continually-evolving preferences of consumers is extremely limited. With data, on the other hand, a CEO can train intelligent software-based algorithms to both (i) present personalized offerings to individual consumers in real-time; and (ii) to evolve these offerings over time as customer preferences evolve. This is but one application of Artificial Intelligence (AI). And data is the fuel that is required to power the AI machine.
When Facebook (NASDAQ: FB) reported its 2Q17 earnings this week, founder and CEO, Mark Zuckerberg, described how Facebook was using AI to help advertisers better target consumers on the platform who were most well-suited to the offering. Historically, marketers would spend an enormous amount of time trying to define the optimal target set. Today, the machine can do it better.
“On the business side, we’re seeing a large shift in the way that marketing works. In the first wave of marketing, people would buy ads and media they thought their customers might watch, like a TV show that had similar demographics, but they wouldn’t know who saw their ads. The Internet gave people the power to target their messages to people who actually might be interested and to measure results much more precisely, and that was a big improvement.
Now you can put a creative message out there, and AI can help you figure out who will be most interested. A lot of the time you don’t even need to target now because AI can do it more precisely and better than we can manually. This makes the ads that you see more relevant for you and more efficient for businesses.” – Mark Zuckerberg (July 26, 2017)
This is a significant innovation that Facebook is uniquely-positioned to deliver. Why? Because Facebook, with more than 2 billion members globally, has the greatest self-updating database of consumer information the world has ever seen. And social data is proving to be the most valuable when it comes to predicting spending behaviour, as research is showing.
Below is how we think about Facebook’s advertising model. Essentially, Facebook can drive the demand for, and supply of, digital advertising on its platform – and other third-party platforms via the Facebook Audience Network.
And Facebook can use AI to enhance both the demand and supply at the same time. Zuckerberg’s example above drives demand by advertisers as his new AI-based consumer-targeting increases the return-on-investment (ROI) of ad spend.
Meanwhile, Facebook is also developing AI to enhance the supply-side of Facebook’s advertising platform. From the same conference call, Zuckerberg said:
“When it comes to News Feed, we currently mostly show you content from people and pages you’re connected to. And we can write this better with algorithm improvements, but the really big improvement from AI will be when we can understand all the other content that’s out there so we can help you discover much more of what matters to you beyond just what your friends are up to.” – Mark Zuckerberg, (July 26, 2017)
More relevant content drives higher membership and engagement – which drives Facebook’s “capacity” to supply digital advertising. And as demand and supply is enhanced by AI, so too are revenues, earnings and shareholder value.
The picture that is starting to emerge is that supernormal value will accrue to those with the best historical, current and future data sources. When it comes to consumer and social preferences, there are really only four “databases” that matter: Facebook (NASDAQ: FB) and Amazon (NASDAQ: AMZN) in the world ex-China; and Tencent (HKEx: 700) and Alibaba (NYSE: BABA) in China.
Montaka owns shares in Facebook, Tencent, Alibaba and Amazon.
Andrew Macken is a Portfolio Manager with Montgomery Global Investment Management. To learn more about Montaka, please call +61 2 8046 5000.