TOP 5 AI TRENDS THAT WILL RULE INDONESIA IN 2020

TREND 1 : Use of AI to personalize messages that drive sales.

Most of us get SMS, Emails, In-app notifications, Web-push from brands that are very generic. Most of the times we ignore these messages and consider them as SPAM. These “one size fits all” messages are not engaging and don’t result in effective sales!

Hyper personalization of digital marketing helps brands stand out and increase conversions with audiences. We see more and more brands collecting valuable insights on individual consumer behaviour and are using AI to deliver highly tailored messages to their customers.

This form of marketing is a paradigm shift from product-centric to customer-centric marketing. Where engaging with the right customers, at the right time, with the right content, through the right channel drives sales.


TREND 2 : Making bots more emotive.

Chatbots are here to stay and they are emerging across major industries in Indonesia. However, many of these chatbots have frequent fallback conditions.

This is because these chatbots don't have the ability to have empathy with their human users. Humans engage with each other by understanding cultural nuances such as basa basi and understanding the various emotions in a conversation.

Chatbots that are trained to interpret human emotions and cultural nuances will outperform standard chatbots. This will reduce fall-backs and significantly improve customer experience and satisfaction scores.



TREND 3 : Use of AI in offline retail stores  

Offline stores are increasingly under pressure due to the strong emergence of Online stores offering attractive promotions.
Also, Loyalty programs run by Offline stores have low membership rates and therefore re-targeting them is challenging. On the other hand Online stores use data to re-target customers and increase their sales.

CCTV’s that are common in Offline stores are primarily used as security devices.
Using EDGE computing, IOT (Raspberry Pi/NVIDIA Jetson) and CCTV images, customer face ID’s are extracted to create a database of visitors who visit stores. This metadata is processed to segment customer purchase history and a customized promotion is created for each customer.
This “In The MOMENT Promotion (IMP)” is pushed via a message to an android APP or POS device of the Store manager, who then can approach the customer with a customised promotion.

TREND 4 : Rise of NLG to create engaging digital content

When it comes to creating digital marketing content for brands, there is a lot of guess work that is done to determine what are the words that can be most impactful in terms of inducing sales and creating a positive impact on the brand.

Graph theory when combined with Sentiment Analytics on customer reviews, chatbot conversations, social media data can help identify Key Themes and Word Correlations.

Natural Language Generation or NLG algorithms then helps to determine sentence recommendations that maximizes positive sentiment & increase sales. So no more guess work is involved!


TREND 5 : Infusing business sense into algorithms

Many CEO's across Indonesia complain of the lack of good analytics within their organisations. This despite having teams of Data scientists within their organizations. It doesn't matter if you have 650 data scientists sitting in another country, crunching algorithms to solve business issues.

Data Scientists, both offshore and locally based ones, need to not just investigate primary data, they need to have the skill-sets to combine data with secondary data assets. Additionally, they need to know the cultural nuances of Indonesia . For example, how do people in Jakarta differ from those in Surabaya. This comes with experience and advice from subject matter experts.

We need to build AI that is less Artificial and more Intelligent! This is what we mean by business sense and this is what algorithms need to have.