How AI and machine learning can be used for marketing: Interview Criteo

ISSUE #67 – From AI to machine learning, data-driven solutions are bringing brands new ways to engage consumers. Alban Villani (pictured below), General Manager of Criteo SEA, HK & TW, shares more about the innovations in marketing, in this innovation spotlight on Artificial Intelligence.


Do you find brands (retail and ecommerce) catching up to today’s changing consumer paradigm?

Consumers want the best of both worlds – they want to browse and buy easily wherever they shop. In response, some brands are starting to offer a mix of online and offline shopping experiences to cater to the rise of the omnichannel customer.

Criteo’s Q1 2018 Global Commerce Review report for Southeast Asia (SEA) showed that omnichannel customers in SEA generate 27 percent of all sales, despite only representing seven percent of all customers. In addition, SEA omnichannel retailers that combine their online and offline data can apply more than four times as much sales data to optimise their marketing efforts. This proves that consistency and connectedness across all touchpoints are now essential for brands.

To achieve that, data analytics comes into play. Retailers and commerce players need to have the abilities to compile and analyse shopper data to better understand and strategise how they should engage customers. Established brands that have built a larger customer base have the resources needed to understand a shopper’s habits by organising and applying data at scale, successfully engaging them with personalised and relevant ads.

Smaller brands and retailers can do the same too. Leading commerce players like Zalora, Lazada, Sendo and Nykaa, are already turning to commerce marketing technology to compete on an equal footing with global commerce giants.



What role does AI and machine learning play in this?

Brands could look at leveraging data-driven solutions to better understand how they can continually engage an ever-growing customer base. AI and machine learning can help brands at both ends of a spectrum – from communicating with their customers to analysing customer insights to improve future experiences. A good use of AI came from Spotify’s billboard ad campaign, where they used their customers’ music streaming preferences to highlight trends such as “Dear person who played ‘Sorry’ 42 times on Valentine’s Day, what did you do?”. The campaign brought consumer data to life in interesting ways and also provided Spotify with insights to better serve its user base.


Spotify crunches user data in fun ways to engage its user base


Moreover, AI and machine learning can help brands navigate complex data collection and storage processes, especially with the recently established policies on privacy like GDPR. Given that AI adoption has it benefits in commerce marketing, we saw it timely to launch the Criteo AI Lab. By combining Criteo’s established engineering expertise in machine learning with R&D in deep learning models, we aim to push the limits in deep learning and AI.


What is one thing marketers need to do in 2018?

There is a concern that the large commerce players will limit access to consumer data that are essential for smaller players. This was seen in Criteo’s study with Forbes Insights, “The Commerce Data Opportunity: How Collaboration Levels the Retail Playing Field”, which saw that 41 percent of retailers were concerned about not having access to information about their own customers and products.

This concern can be addressed with collaboration. When brands value the importance of collaborating and pooling data assets, they can help meet customer needs and drive value for their businesses. In fact, 71 percent of retailers surveyed for Criteo’s study with Forbes Insights mentioned that they would contribute online search data to a pool. Sixty percent were already part of a data cooperative, while almost 70 percent have indicated that they were satisfied with their collaborations as well as the data they receive.

Being built on the open internet, we strive to break down walled gardens and make data sharing more accessible for all. With these data on hand, brands can identify shopper behaviourial trends, strategise in their best interest, and drive engagement and sales. Data collaboration is also important in helping to level the playing field against the giants.

However, as data privacy and protection become top of mind, brands must be conscious and committed in ensuring that customer privacy is upheld with utmost importance. Businesses should therefore only consider data collaboratives that work with non-directly identifiable information, or pseudonymous data as stated under the GDPR. This offers more warranties to shoppers in terms of confidentiality.

As consumers continue to evolve, the retail landscape cannot stop either. Brands need to stay up to date with the times to know how to navigate the complexities of engaging shoppers.



Interview with Alban Villani, General Manager, Criteo SEA, HK & TW

Published in FOCUS Magazine — Issue #3 2018 “The Innovation Issue