Will Chat GPT and other Smart AI transform Email Acquisition and Digital Marketing or cause more issues?

Will Chat GPT and other Smart AI transform Email Acquisition and Digital Marketing or cause more issues?

There is a lot of hype around ChatGPT but if you read Scary Smart or followed some of the latest conversations on it, Smart AI has many risks. So when it comes to email what are the benefits:

Smart AI can help improve email marketing by generating personalised and relevant content for each recipient. This can be done by analysing your own CRM data on customer behaviour, preferences, and past interactions with the brand or leveraging third party insights, like @esbconnect.

Smart AI can also help automate certain aspects of email marketing, such as creating subject lines and crafting email copy. Smart AI can generate compelling and persuasive language that resonates with the recipient by utilising natural language and machine learning techniques. Powering it with your data will ensure relevancy.

In addition, Smart AI can help increase engagement by creating interactive and dynamic content, such as quizzes or polls, that encourages recipients to interact with the email and the brand.

Overall, Smart AI has the potential to significantly improve the effectiveness of email marketing by generating personalised, relevant, and engaging content that resonates with recipients.

However, there are risks on a basic level:

First, there is a risk of over-reliance on technology, which could lead to a lack of creativity and originality in the marketing content. While Smart AI?can generate high-quality language, it relies on past information but over time, your content may become repetitive.

Second, there is a risk of generating irrelevant or inappropriate content if the data fed into the model is not accurate or up-to-date. Smart AI relies heavily on the quality of the input data. If the data is incomplete or biased, the generated content may not be effective or could even be harmful.

Third, there is a risk of privacy violations if sensitive data is used to train the model or if the generated content contains confidential information. Companies must ensure that they follow best practices for data privacy and security to protect their customers' information.

Fourth, there is a risk of creating a negative user experience if the automated messages are not personalised or if the recipient feels like they are receiving spam e.g. you ask them the Smart AI to send when they are most likely to convert and it sends them 200 emails at 5pm on Friday.

But the scary part, is we cannot forsee every outcome of our instructions.

Consider a simple scenario: You are a supermarket brand and want to boost your email program's revenues. You use Smart AI to send an email with personalised content to each customer when they are most likely to convert. Customer A, is trying to give up alcohol, but the highest ROI seen from them is on alcohol emails sent to them late at night. So the AI starts to create deals and content, targeting them late at night and encouraging them to buy more alcohol. It works. However, it results in a complaint from the customer, requesting you not to send alcohol deals to them, as they are trying to give up. You program the AI to exclude sending anything that contains alcohol to them. However, the Smart AI is still working to maximise your revenue, so it sends them non-alcoholic versions of their favourite tipple but ensures they are taken to a page where they can see both options. Thus assuring the customer is still tempted and returns a high revenue. To avoid this, you now say do not send them to the 'drink section' on the site. However, the Smart AI now understand the individual has an addictive personality and looks for other products that deliver similar temptation and addiction. Although some products may limit how much can be sold in a single transaction, the Smart AI can send them multiple emails to make repetitive purchases.

The issue is that smart AI will always look for the most efficient way to achieve its goal and will be more intelligent than us. We cannot foresee every likely outcome of instructions, and our instructions are full of nuances. Smart AI poses many risks; you must consider the negative consequences before implementing it.

Are you scared or excited about the future of Smart AI?