Google’s success originated in one simple insight from its founders who realized that the sprawling, chaotic mass of material that was cascading onto the world wide web could be tamed by ranking search results according to their popularity.
They have not looked back since, making millions from online advertising off the back of the algorithms that power their search engine, even as they direct and nudge how consumers interact with the world. The use of algorithms in search is just the tip of the iceberg today. Many other platforms are already using machine learning (ML) algorithms to enable AI-based outcomes, fueling the virtuous cycle of meeting consumer needs and shifting consumer behaviors.
The primary goal is to understand every consumer’s unique buying behavior and psyche, and surreptitiously nudge them to click on that ‘add to cart’ button.
Take for example Bytedance’s flagship products, TikTok (known as Douyin in China) and Toutiao. TikTok uses AI in various ways, from facial recognition for their filters through to the recommendation engine in the “For You” feed. The platform is capable of understanding and analyzing text, images and videos using natural language processing and computer vision technology, enabling them to serve users with the content that they find most interesting, and empower creators to share moments that matter in everyday life to a global audience.
Similarly, for Toutiao, which acts as a one-stop shop for aggregated Chinese news, its unique selling point is gathering stories from thousands of outlets and using AI to personalize content according to the individual user’s interests and preferences, and in so doing, increase user engagement and time spent on the platform.
The use of predictive algorithms for content recommendation and content discovery based on a user’s viewing history and patterns is also one of the tricks used by over the top (OTT) content providers to foster loyalty and increase content consumption.
In a region with a highly fragmented OTT landscape, that is a result of differing content preferences, a myriad of providers and distributors and the rise of many innovative video streaming apps, it is no longer a nice-to-have feature but a must-have for any successful OTT platform. The algorithms used will play a very important role in driving all data analysis, predictions and trends for OTT businesses.
For iQiyi, one of China’s largest online streaming platforms, it was the use of big data and algorithms that led to the decision to back ‘The Story of Yanxi Palace’, one of the biggest blockbusters a couple of years back. The algorithms helped conclude that the show, set in the court of Emperor Qianlong during the eighteenth century, would likely be a winner, based on analysis of past viewership patterns of similarly themed shows.
To embrace them, brands need to be armed with a wide variety of content for consumers to discover, engage, remix and re-engage with. Content also needs to play to the different strengths and opportunities that each platform offers as they evolve.
Big data analysis was also deployed to determine the optimal time and pace of releasing new episodes to build up viewer anticipation. In spite of the absence of A-listers in the cast, the drama series about back-stabbing concubines garnered more than 13 billion views by the end of its run on the platform, becoming also a hit in Hong Kong and the South East Asian region. Though mostly fictional, together with several other popular imperial period dramas, the show ignited much interest in Chinese history and culture with a global audience.
From global giants like Amazon to local online retailers like Snapdeal and Flipkart, and startups like Zomato, the ubiquitous algorithm would likely be one of the most important tools of the trade in a market as diverse as India. For an online business like Snapdeal, one of India’s largest online marketplace, 35%-40% of sales on the platform are driven by its algorithms, as reported in an article by The Economic Times. No two people logging onto the shopping portal will see the same home page; instead, what they see will be determined by algorithms in the back-end processing data on what they like, things they plan to buy, what their friends or even people with similar profiles might have bought.
The primary goal is to understand every consumer’s unique buying behavior and psyche, and surreptitiously nudge them to click on that ‘add to cart’ button. Even though current ecommerce penetration in India only stands at 28% according to a report by Statista, it holds out a huge promise of a major shift in consumer shopping behaviors guided by the use of algorithms to help navigate the market’s diversity in geography, culture, language and traditions.
With every major tech company and even smaller start-ups in this region investing in and building the future of their businesses on artificial intelligence, even going beyond their platforms and getting involved in smart city infrastructures as well as connected homes and cars, a culture driven by algorithms seems to be an inevitable near future.
To embrace them, brands need to be armed with a wide variety of content for consumers to discover, engage, remix and re-engage with. Content also needs to play to the different strengths and opportunities that each platform offers as they evolve. And whilst algorithms help ingest and analyze huge amounts of data and in real time, they are ultimately built on a set of rules to be followed, and if used as a crutch by marketers, will instead impede true creativity and innovations.
This is the fourth in a five-part series written by Sharon Soh, Head Of Strategy APAC at IPG Mediabrand’s UM, that delves into how platform development will shape the way brands connect with consumers across Asia over the next decade.