Not only human: how to effectively transform digital experiences with AI
There’s an art to artificial intelligence (AI). Scott Ewings of TPXImpact urges marketers to provide AI with robust training before releasing it into the world.
Automation shouldn’t come at the expense of personalization, says TPXImpact’s Scott Ewings / Andriyko Podilny
2023 has been widely referred to as the “breakout year for AI”, and businesses across all industries are capitalizing off this buzz of excitement by dabbling in the likes of ChatGPT and AI image generators, drastically changing the interaction between these brands and their customers, and transforming their businesses entirely. While new business opportunities generated by AI solutions may add up to $15.7tn to the global economy by 2030, how can marketers best take advantage?
While Open AI’s ChatGPT reached more than one million users within five days of launching, the launch of Google’s Bard AI service wiped $100m off Google’s share price, showing how stark the difference between success and failure can be.
AI services can now write emails, blog posts and even code within a matter of seconds, making it an extraordinarily powerful tool if embraced in the right way, but brands need to understand the technology and test use cases, rather than betting the house on its capabilities.
Garbage in, garbage out
These services can be especially effective if they are used to complement – rather than replace – interactions between customers and human company representatives. TPXimpact demonstrated this when we partnered with the Cyber Helpline to deliver free support to UK cybercrime victims via a chatbot. This was especially important during the pandemic when access to call centers was limited due to self-isolation rules.
At the same time, the shift to online behaviors also led to a proliferation of cyber-attacks too numerous to be dealt with manually. Cybercrime victims were able to access useful advice using the newly developed chatbot, while human volunteers were able to concentrate on more complex and high-demand cases.
But there are other ways that AI can be used more specific to the marketing industry today. For example, as a creative thinking accelerant: developing initial visual concepts, first drafts of social media copy, or developing simple exemplar code. AI solutions can also help guide the decision-making processes of digital strategists, by delivering insights from datasets too dense for anyone other than data scientists to interpret. In this early stage of the adoption cycle, the primary skill to develop is effective prompting for these services. ‘Garbage in, garbage out’ is the general rule here.
Segmentation to transformation
Customer segmentation is something that AI excels at which can exponentially increase the effectiveness of marketing campaigns. AI can be transformational at both stages of this process, segmenting between a running shoe customer who is primarily looking for style, for example, versus one who is looking for better running performance. It can then create and deliver bespoke campaigns to each segment.
This can contain more targeted and accurate recommendations of products. All this serves to quickly progress the customer from awareness to action, while AI can also then automatically integrate purchasing information into future campaigns.
In this context, it is also easier to see how AI can help marketing teams with delivering a better brand experience and effectively playback information to the customer based on their unique history. 62% of customers they surveyed said they were open to companies using AI if it can significantly improve processes and the level of service they receive.
This is because, when interacting with a brand, 70% of customers say that trouble-free transfer of their information between individual channels and departments, including adapting from previous interactions, is highly influential over their trust in that brand.
There are also productivity gains that marketing teams can benefit from in terms of AI enabling them to better manage multiple campaigns at the same time, something especially important within agencies.
Humans, take note
While the introduction of new AI technology comes with much excitement, there are also some ethical concerns and potential risks presented by the proliferation of these new technologies. AI models can propagate unfair biases if they are trained with the wrong kind of datasets or inherit these directly from their human trainers.
As well as giving a bad experience to end users this also can severely damage the reputation of companies – something the marketing industry is clearly more sensitive to. This is why it is even more important that AI goes through robust training and comprehensive datasets before thinking of releasing them to the wider world.
For marketers considering AI as a tool, there is clearly a balance that needs to be struck between the productive and the personal which has a direct impact on determining effectiveness. AI should be used to enhance good marketing, but automation shouldn’t come at the expense of personalization or, in the worst case, cause lasting damage to a brand’s reputation. While AI can be a smart way for marketing to deliver a better customer experience, we still shouldn’t lose sight of the human element.
Content by The Drum Network member:
Manifesto, the multi-award winning digital experience agency has joined forces to represent the Digital Experience capability of TPXimpact.Find out more