Can Machine Learning Solve the ROI Problem?

Artificial intelligence (AI) isn’t just the buzzword of the moment-it holds the potential to significantly change the marketing landscape in China and help companies market to the segment of one. eMarketer’s David Green spoke with William Bao Bean, managing director of startup accelerator Chinaccelerator, about how machine learning and AI are already helping marketers in China with their mobile messaging strategies, as well as budget allocations for influencer marketing.

William Bao Bean, Managing Director Chinaccelerator

eMarketer: What’s an area where AI is helping companies in China improve their marketing efficiency?

Bean: One of our companies, OCheng, is applying AI to WeChat commerce. WeChat service accounts are allowed to send a message to their followers once a week. OCheng can take customer relationship management data, look at the content people are engaging with, their purchase history, etc., and use machine learning to personalize these weekly messages. Instead of sending one message to a million people, you can send a million messages that are unique to each user every week.

eMarketer: Is this process entirely automated?

Bean: The content has to be written around the products the marketer is pushing that week. But the way the messages are written and the offers that are being shown to each person are personalized through automation. And when users write back, an AI-assisted chatbot can answer their questions.

You not only have personalized content going to each user, but you can also have one-on-one conversations with WeChat users without a human, or you can have a chatbot and human working together.

eMarketer: What types of brands are working with the tech?

Bean: The Richemont Group, Dunhill, Macallan and Sephora are some of the brands that are leveraging this type of technology to drive higher sales. It’s not some pie-in-the-sky platform for them-it’s about dollars and cents, and there’s real ROI return on investment.

eMarketer: Is it easier to achieve this on WeChat than Facebook?

Bean: You can do a lot more within WeChat because it’s a more open platform than Facebook. A lot of the techniques you use on WeChat you can also use on Facebook. Facebook should start catching up, but the difference is how the two platforms make money.

WeChat makes money on transactions, whereas Facebook makes money on ads. Facebook wants to capture ad revenues, whereas WeChat parent company Tencent wants to capture the commerce aspect. Marketing is important, but the opportunity to capture a percentage of the transaction as opposed to a percentage of the marketing budget is a bigger opportunity over time.

eMarketer: In what other areas is machine learning being deployed to improve marketing outcomes?

Bean: Influencer marketing is big in China. Influencer platform Robin8 uses machine learning and natural language processing to analyze the content that key opinion leaders KOLs push out, and then analyze the impact-who’s engaging with that content.

Robin8 can do end-to-end attribution from KOLs all the way through to purchases on WeChat. This gives marketers an idea of what kind of content from which KOL works with different types of users, and they can optimize their KOL buy.

Leave a Reply

Your email address will not be published. Required fields are marked *