Dinnertime conversations: At home with big brands

The ‘Nestlé Xiao AI’ is a product of the Beijing Linglong Company, a joint venture between voice-recognition specialist Iflytek and e-commerce giant JD.com. It offers recipes specifically developed by Nestlé, answers nutritional questions for eating well (with Nestlé products of course), and even plays songs. To activate, users say “Nestlé Nestlé” in Mandarin, in yet another exercise of brand reinforcement. The speaker retails for an affordable RMB 299 (about US$45). Or, it could be given free to consumers who buy enough Nestlé milk powder in bulk (with built-in voice commands to have even more milk powder easily delivered). 

Campaign spoke with Nini Chiang, CMO of Nestlé China, about the product.

Campaign: What is Nestlé's rationale behind having a branded speaker?

Chiang: This is our first trial enabling us to directly enter consumers’ living rooms via ‘Nestlé Xiao AI’. We realised in the new digital ecosystem, food and beverage manufacturers cannot provide just products, but with more valuable information that can help consumers better their lives. I asked Ogilvy, our advertising agency, to share new trends in the market and got to know about voice-AI.

Working with Linglong was a very good starting point even though they are a small startup. After a few rounds of discussions, we decided to make 2,000 tailor-made speakers for Nestlé.

Nini Chaing

Nestlé has a worldwide mission: to enhance quality of life and contribute to a healthier future. We have a lot of good content related to nutrition science, which we provided Linglong to train their AI. We started with FAQs from our consumer hotline, such as “is it normal for coffee to taste sour?” Now, the complicated part is every question may have many potential answers, so it was very labour-intensive for us in the first four months. 

There is a lot of ‘pseudoscience’ floating around in China, and we believe this product can offer more accurate information than those found through online question-and-answer websites. In fact, our legal team has reviewed every answer to make sure every single piece of information ‘spoken’ by ‘Nestlé Xiao AI’ is credible. 

Campaign: How is this trial increasing Nestlé’s brand equity and how are you measuring it?

Chiang: It’s too early to say, but if you ask me for particular marketing objectives, one is to strengthen our corporate image; two is to gather more interactive consumer insights compared with focus groups and surveys. Regarding data privacy, we made sure we collected data in a compliant and useful way, like ‘what kind of age groups asking what different types of questions’.

Globally, consumers have mixed feelings about smart devices being ‘too smart’, like smart TVs ‘listening’ and recording conversations—very scary. I don’t have a full picture of consumer attitudes about this, but in China, we use Alipay, WeChat and wi-fi everyday, everywhere. In reality, what you bought last week or where you went today can be tracked. There are two reactions: some consumers are not too conscious about it, or believe this makes their lives easier since big data allows marketers to have enough knowledge of them to send personalised brand messages; while some feel more worried than others.

Kitchen companion: The Nestlé Xiao AI aims to talk with families about cooking tips and nutritional information in their homes.

Campaign: How much did you spend on this voice-AI project?

Chiang: Not a large amount of money. Basically, we co-worked with Linglong. We provide content for free while they made the machines. It’s a collaboration.

Our next steps are to increase the number of question-answer combinations. We need to enhance the content so as to drive more household penetration for other Nestlé business units.

Campaign: What is the biggest lesson for you in this first foray into AI?

Chiang: We wanted to learn about how to work with a hi-tech company. In order to do that, we need to find a way to talk to them so that both parties can understand each other; so people in marketing can speak the same language as technical guys who are focusing on quality of their tech and accuracy of their AI tech. It didn’t occur to them they could explore different financial models, like selling media through their machines. On the other hand, we, as a so-called traditional food manufacturer, learnt so much about AI.