Beware the bots – the rise of the machine Asia style

bots Asia market
Chat bot with AI helping of users speakers of different languages. Image by Andrey Suslov | Bigstockphoto

Organisations worldwide are building a digital workforce of bots to alleviate pressure on their call centres, speed the resolution of simple enquiries and lower the cost of support. But while North American, European and Chinese firms are building, investing in and deploying bots, large parts of Asia-Pacific has lagged behind in the bot revolution. The region is fast catching up, however, with $1 trillion of value at stake between 2022 and 2030. In this article, I take a closer look at the fast-growing bot market in Asia.

Language, economics and digitalisation slowed adoption in Asia

There are a number of reasons why the bot market didn’t take off as fast in Asia as in other world markets. One issue is that bots substitute for people, so in markets where labour is scarce or expensive (such as the US or Europe) the incentive to adopt bots was higher. Another key barrier has been language. The adoption of chatbots was initially driven by US start-ups, which meant Anglophone countries (including Australia and New Zealand) were in the first wave of adopters.

Bots have struggled to understand English dialects, so with its wide variety of languages, accents, dialects and slang, it’s not surprising that Asia-Pacific has presented even more of a linguistic challenge.

Bot adoption is speeding up in Asia

After a slow start, there are signs that bots are now taking off in Malaysia, Vietnam, Indonesia, Singapore and the Philippines. Partly stimulated by pandemic-induced digital acceleration, everyone from healthcare providers to financial institutions, transport & travel firms, education providers, retailers and human resources are now looking at how they can make use of bots to automate manual processes and digitalise their offerings.

According to the 2021 Asia Chatbot Trend Report, which surveyed customers in Hong Kong, Singapore, India, Japan and South Korea, 66% of those surveyed had heard about chatbots and, of those that had, 85% said they’d used chatbot services in the previous six months.

Respondents were of the opinion that this was the “perfect time” to introduce chatbots, and believed they are helpers of humans not competitors for their jobs. The report was put together by Makebot, whose AI-powered chatbots can link to a range of social networking sites such as Facebook, WhatsApp and Line. Its bots are multi-lingual and last year the company provided support for the K-pop entertainment platform Weverse as well as Korean Air.

Chinese customers see chatbots as ‘dear friends’

China has been one of the exceptions in Asia, with the tech-savvy nation being an early adopter of bots. Back in 2014, Microsoft launched a chatbot called Xiaoice which chats and jokes with customers who see the bot as ‘a dear friend’. By 2018, Xiaoice was chatting more naturally and designing fabrics in partnership with the China Textile Information Center and the China Textile Development Center. A major advance came when ‘full-duplex voice sense’ was added, meaning it could communicate in a more natural way by listening and talking at the same time.

Harry Shum, Executive Vice President, Artificial Intelligence and Research at Microsoft said: “We’ve been on a journey with conversational AI, emotional intelligence and chatbots that I’m not sure anyone else in the industry has undertaken yet. It’s inspiring to see this coming to life today in China – perhaps more than any country in the world – with our advances in AI and perhaps most visibly with our extremely popular social chatbot, Xiaoice, who has had over 30 billion conversations with 100 million friends.”

By July 2021, Xiaoice’s value reached $1 billion and it was spun off from Microsoft.

Vietnam looks to bots to boost tourism

In late February 2022, Da Nang, a tourism hub in Vietnam, became the second city in Southeast Asia to deploy chatbot tech to support tourists. During the six-month trial, the chatbot provided a range of information in Vietnamese and English about destinations, events, weather and even ATM locations, as well as e-coupon promotions for attractions. Its countdown feature reminds tourists about key events and even helps them plan routes to get there. In future, the bot will add Chinese, Korean and Japanese support.

South Korea found K-bot was not quite K-pop

With South Korea identifying AI as core to its future (launching its AI National Strategy in December 2019), the country wants to become as well known for its bots as it is for its bops. This meant there were high hopes for Lee Luda (a homonym for ‘realized’ in Korean), an open-domain conversational AI chatbot developed by ScatterLab and launched in late 2020. The firm had raised over $5.9 million and was backed by giants such as NC Soft and Softbank.

Luda used deep learning and over 10 billion Korean language datasets to create a realistic bot that was integrated into Facebook Messenger, and which simulated a 20-year-old female college student. Users were encouraged to converse with Luda, but the result was not what the developers had anticipated. Shortly after it launched it became apparent that users were training Luda to spew hate speech against women, sexual minorities, foreigners, and people with disabilities. It was also sharing private details, including bank account numbers, taken from its dataset. ScatterLab was forced to suspend Luda on January 11 – a mere 20 days after she was launched.

ScatterLab was by no means the only company to face such challenges – a similar fate befell Microsoft’s Tay, when users began training it to make offensive remarks. But the Luda incident has provoked a national conversation about AI ethics in South Korea, which will stand the country in good stead as it further embraces AI.

Singapore’s StarHub uses bots for seamless customer service

StarHub had all the usual telco support challenges – siloed systems, manual and archaic processes, and frustrated agents having to toggle between platforms to resolve customer queries. Understandably, its customers wanted more choice of channel combined with a more efficient and digital experience. So StarHub turned to Reliance-owned Haptik to build a chatbot for its Facebook Page to support its digital TV, cable, IPTV, broadband, home phone and mobile businesses. The new chatbot can answer queries around prepaid plans, fibre broadband, new offers and contracts.

The key to bot success is data

It’s important to see chatbots as another, and very useful tool, in the customer support mix in Asia-Pacific. But while vendors are working to add empathy and improve them linguistically, they still have their limitations and are not a cure-all. Microsoft admitted that it focused on empathy rather than intelligence when it rolled out Xiaoice, for example. Putting ‘the brains’ into bots requires bots to have access to good quality, real-time data. If bots don’t have the data to resolve queries and can’t seamlessly transfer customers with complex queries to agents – something that the Haptik bot used by StarHub can do – CX can drop, rather than increase.

Fredrick Edwall, EVP Sales & Marketing, from Sweden’s Subtonomy, which provides accurate, real-time data on network performance for customer support in the telecoms industry, explains: “Chatbots have a vital role to play and their capabilities are developing fast. But what’s often overlooked is that they can only ever be as good as the data provided to them. So unless you feed them with up-to-date data about what’s happening on the network, for example, you stymie their ability to mop up simple queries. Our platform provides the same real-time data about network performance to all support channels – call centre agents, self-service and bots. Ensuring that all channels are working from the same, accurate, real-time data provides the essential foundation for omnichannel customer support success.”

Investors pouring money into Asian start-ups

Not only are Asian customers becoming bot users, but they’re also increasingly becoming bot builders, with investors keen to support them.

Take Singapore-based AI Rudder as an example. It recently announced it had closed a $50 million Series B round led by Tiger Global and Coatue. AI Rudder’s platform provides chatbots for banking and finance firms and supports 20 languages as well as dialects such as Singlish. “The ubiquity of technology and SaaS has allowed us to impact businesses across continents – in countries like Mexico, Kenya, Australia and beyond,” commented Kun Wu, AI Rudder’s MD. “We’re excited to lead this next phase of growth into new territories and industries to supercharge omnichannel customer experiences for our clients.” The new investment round comes just four months after AI Rudder closed a Sequoia Capital India-led $10 million round.

And it’s not alone.

It may have taken India’s Uniphore 15 years to get there, but its latest Series E investment round of $400 million values it at $2.5 billion, making it a chatbot unicorn. The company, which provides conversational AI, also recently entered into an agreement with VIS Global to provide Asia-Pacific organisations with comprehensive communication and customer experience capabilities.

Meanwhile, India’s inFeedo just received $12 million in Series A funding led by Jungle Ventures, with participation from Tiger Global and returning investors like Bling Capital. Amber, inFeedo’s chatbot, supports a wide range of customers (including Samsung, Xiaomi, Lenovo, TATA, Godrej, Bharti, Unacademy, Paytm, OYO, Lenovo, JD.ID., Tiket.com, Mediacom, Sunlife, BukuWarung and Aboitiz) across Singapore, Malaysia, Indonesia and the Philippines. Rather than focusing on listening to customers, Amber listens to employee sentiment – helping customers boost employee experience (EX).

$1 trillion bot-ortunity

A study by Kearney points to significant country differences across Asia, with highly digitalised and efficient economies such as Singapore leading the charge, but it concludes that the use of AI and bots in Southeast Asia alone will see a 10 -18% GDP uplift by 2030, equivalent to nearly $1 trillion in added value. The authors say: “The value of AI is not in the technology itself, but in how it can solve business problems and deliver value in the near term”. This is music to Omnisperience’s ears. And in our view, as the Asian Tigers wake up to the potential of bots, in the next 10 years we’re going to hear them roar.

Related article: Scammers turn to bots and automation to stay undetected

Be the first to comment

What do you think?

This site uses Akismet to reduce spam. Learn how your comment data is processed.