The Social Commerce Love Affair With AI

Artificial intelligence (AI) promises to change the world – but will it change commerce?

You’d be forgiven for thinking the world’s gone AI crazy these past few weeks. Indeed it has, in a way. At the end of last month, Microsoft said it would invest $10 billion in ChatGPT firm OpenAI.  

OpenAI, for its part, has launched a premium version of ChatGPT.

And amid a slew of announcements from tech giants and smaller players alike heralding the use of AI in every social, commercial and academic setting imaginable, there’s the evergreen ethical debate of how AI might be tied to privacy concerns, bias and corporate responsibility.  

Most immediately, some of the biggest names in tech are finding an early roadmap in using AI to streamline and personalize commerce – at the consumer-facing level (to boost sales conversions) and to enable businesses to better understand and target those customers.   

The data is in place, of course. We note that there’s the potential (no matter which company is using AI) to give a 360-degree view of a consumer across all channels as they interact in an omnichannel fashion and receive personalized offers and promotions in real time.  

Chatbots and smarter virtual shopping assistants can keep users engaged in and out of social media channels and apps. At the same time, by leveraging natural language processing to give search results a more “human” and conversational rendering that goes beyond simple rankings, which has the potential to change commerce in infinite ways.

Moving Beyond Search

But tech firms’ ambitions stretch well beyond search. During this past week’s earnings call, Alphabet’s CEO Sundar Pichai said that the firm will aim to “unlock the incredible opportunities AI enables,” adding that he felt the technology was moving towards an “inflection point.”

“[AI] is a powerful enabler for businesses and organizations of all sizes,” Pichai said. As for how businesses might harness AI, CBO Philipp Schindler stated that AI is already being harnessed to create “smart bidding, which uses AI to predict future ad conversions and their value, which in turn help businesses stay agile and responsive to rapid shifts in demand.”  

As we noted in our own earnings coverage on Alphabet’s earnings, AI-powered campaigns such as Performance Max (PMax, which automates targeting and delivery), have improved conversions by about 12% for advertisers.

Elsewhere, and as detailed in PYMNTS’ coverage of Meta Platform’s earnings this week, the company’s longer-term AI strategy is designed in part to have “more relevant content recommended by our AI systems,” in the words of CEO Mark Zuckerberg, rather than having a model in place where people “follow’ accounts.

The company’s “monetization efficiency,” as Zuckerberg put it, has doubled for Facebook in the past six months as AI continues to make inroads in the company’s key advertising business. Advertisers saw 20% more conversions than in the year before, according to call commentary.

The company sees longer-term opportunities to scale conversions via Shop Ads and click-to-message ads (Zuckerberg said click to message is seeing a $10 billion run rate), which link consumers and businesses into direct conversations across WhatsApp, Messenger or Instagram.

During the question-and-answer session with analysts, Javier Olivan, Meta COO, said that they are continuing to test Shop Ads and said, “we’re seeing increased performance by helping direct the consumer … to where they’re most likely to purchase…it’s a small base, but just to give you a sense, we saw triple-digit growth in both revenue and adoption across Q4.”  

That last comment is telling. The “base” is small, and growth off a small base has been significant. But it’s still relatively early days yet for AI, and there are still hurdles and hiccups.

As reported here, retailers are thus far mixed in their assessment of AI’s usability and are mindful of ChatGPT’s limitations. ChatGPT is more likely to give an answer — right or wrong — with total confidence, whereas other chatbots are designed to tell users “I don’t know” when they can’t answer a request. Even OpenAI CEO Sam Altman has cautioned against relying on ChatGPT “for anything important right now.” On Twitter last month, Altman said the tool was “a preview of progress; we have lots of work to do on robustness and truthfulness.”

The promise of AI is somewhere just over the horizon, at least at the moment. Getting there, and making money from it, will involve both trial and error, for Meta, Microsoft and Alphabet — and everyone else, too.