AI-powered marketing for optimum performance

AI has transformed the world of marketing and is expected to drive enormous changes. Discover how AI enhances customer relationships by generating automated decisions based on data analytics, observing audience behaviour and economic trends. Learn what are the best AI marketing tools to optimise your campaign performance.  

Question & Answer

For the Mike, the AI agent, have you conducted any test to examine how different tones used by AI agents affect the audience’s perception and behaviours?

[Tim]
The connection between emotion and audience behaviour is part of our larger Audience Behaviour study, because different behaviour can be triggered by different attributes and speaking emotion/style is just one of them. We are also exploring to tap on AI to suggest different speaking styles/emotions when generating scripts for different targets. 

For Mike, how long does it take to train him to sound like someone famous like Ironman and how much would that cost? 

[Tim] 
There are different types of technologies for different use cases, from Open Source models to commercial tools. It doesn’t cost much to do AI clone/synthesis technically, however, AI adoption is a long term investment rather than a one-off exercise. You have to consider other operational costs on how you can integrate AI into your existing business process, how to run on-going support and eventually how to transform it into a sustainable business case.

As a follow up to the guidelines/ethics question, what do you think about 'transparency' – i.e., do you feel brands should indicate when output is created by AI or with input of AI? If yes, for which cases and when not?

[Tim] 
It will depend on the actual use cases, where and how AI is used, as well as whether it is aligned with your organisation’s AI governance guidelines/policies. The key thing is to understand what role AI is playing and how it will affect the audiences’ behaviours or decisions, and how that could potentially affect your business. However, one suggestion is always have a human-in-the-loop approach when you using AI.