Artificial intelligence (AI) has become one of the most talked-about topics in marketing lately.
Depending on who is speaking, AI’s potential impact on marketing can vary dramatically. To some, AI is a tool which will help marketers improve performance, to others it’s the technology which will eventually replace us.
But what do marketers think about AI now? Are they integrating it into their daily marketing activities – or are they still trying to figure it out?
To find out, Econsultancy, in association with Oracle, recently held roundtable discussions with dozens of client-side marketers in Manila. At a table hosted by Philippe Soriano, CEO, Mediafied, attendees shared their views, experiences and challenges regarding using AI in marketing.
The discussions are summarized below but before we start, we’d like to let you know about an upcoming event in Indonesia. On 20th June 2019 from 9AM-1PM, Econsultancy, in association with Oracle will be hosting a free roundtable event for client-side marketers at the Shangri-La Hotel in Jakarta.
For more information and to book your spot, please visit the event site.
The first thing participants agreed is that AI is not yet being used by marketers at most companies. This was not due to a lack of interest but instead because many attendees felt that they were not yet prepared to integrate AI with their existing marketing technology. Throughout the day, delegates arrived at four main reasons why AI had not yet been adopted by marketers.
1) Marketers do not know what AI can do for them
Marketers, one participant explained, will most likely use AI through an algorithm which interprets historical data and makes predictions about the future. What most attendees did not yet understand, though, is how to apply that application of AI to their business.
The challenge for marketers, said another, is that AI will be used in different ways for different industries.
For example, in ecommerce, AI could be used to analyze an individual’s past purchases and make dynamic product suggestions or offers. Similarly, a brand with a large content marketing programme could use AI to make personalized reading recommendations based on a customer’s viewing history. And B2B firms might use AI to review data of existing customers and determine which new leads are likely to become high-value customers.
15 examples of AI in marketing
Another attendee pointed out that AI could also be used by marketers for general tasks such as improving copy, optimizing ad delivery, or as the engine behind chatbots for customer support.
Regardless, delegates agreed, marketers need to understand more about what AI can do before they will know how to use it for their business.
2) To be effective, AI requires data which marketers do not have
For every AI application, one participant explained, marketers need to have large quantities of data to train the algorithm. While there is no agreement on how much data is enough, the more data marketers use to train the AI algorithm, the more likely it will make accurate predictions.
For example, a B2C ecommerce site which has thousands of transactions per day will be more likely to benefit from AI than a B2B one which may not have that many in a month.
Most attendees felt that while their companies did generate a lot of data, they either did not have access to it or the data was not yet in a format which could be used to train an AI algorithm.
3) Marketers lack the skills to manage AI projects
In addition to data, adopting AI technology for marketing requires knowledge across many disciplines which participants admitted was lacking in their teams.
First off, acquiring and managing data for AI is challenging. Data volumes required for AI are typically larger than what marketers normally handle and as data comes in many formats, it needs to be ‘cleaned’ before it can be used. Marketers, therefore, need to improve their data management skills before starting AI projects.
Additionally, said one participant, there are many kinds of AI algorithms and unless marketers understand what they do and how to optimize them, they may end up using them incorrectly and gain little value from them.
Finally, said another, using AI with legacy marketing technology would not be easy. Participants agreed that AI would need to be integrated with existing marketing systems, but most were unsure how to do so.
4) Marketers need to better understand how AI adds value to the business
A final point raised by delegates is that before dedicating time and resources toward AI, marketers need to know more about how it will deliver return on the investment.
For example, how many new sales will product recommendations deliver? How much will AI-generated lead scores improve sales? How will marketers know whether AI ad optimization produces better results?
Participants understood that the answers to these questions will be different for each company, but without any idea of what to expect, it was unlikely that AI projects will get off the ground. So, while delegates were eager to learn more about AI and how it can improve their effectiveness, they agreed that they had a lot to learn before AI becomes a part of their marketing programmes.
A word of thanks
Econsultancy would like to thank Philippe Soriano, CEO, Mediafied for hosting the Harnessing Technology table and Oracle for sponsoring the event.
We’d also like to thank all the marketers who participated on the day and offered their insights about AI and marketing. We hope to see you all at future Econsultancy events!