AI in advertising. AI in marketing. AI in everything. These are the subject lines and headlines of the day, and I must freely admit I am just as guilty as everyone else of using them to garner attention. Still, it’s worth a double click into them to see what role AI can truly play for you in your day-to-day media.
Every adtech and martech company has quickly pivoted to leveraging AI and machine learning in their storytelling, regardless of how much that is true. This is not new, and most of these use cases are barely touching AI. They are mostly algorithmic in nature, and rules-based machine language at best.
When a company starts to leverage ‘Generative AI’ in its messaging, there is more likely some real AI involved, but even that can be confusing because it may be used in a way disconnected from the core offering they provide.
No rules and regulations
There are no rules and regulations about how you can leverage these terms, and there’s no governing body to hold these companies accountable to how they market themselves in this regard. The only governance is around false or misleading statements.
Many brands are using Generative AI in image creation for their ads. Brands like Coca Cola have been leading this charge, and doing it well. This is an easy use case because it simply reduces the time required, and in many cases is essentially crowd-sourcing artistry through the web and a combination of AI and user-generated inputs.
This creative use case for AI can be paired with an art director to come up with new and unique ways to present a brand, and it can be done very efficiently.
Don’t use as a stand-alone
Generative AI can also be used for copywriting, but my recommendation is to not use it as a stand-alone. It should be used as a writing partner for a copywriter. It can offer a first draft, an editing solution and an idea generator, but it should not be your only source of copywriting.
You should also keep in mind that whatever copy you put into it is now open-sourced and available to everyone else. If you are trying to keep something close prior to a launch, I don’t recommend you have an AI tool do your writing for you.
In paid media, there are a number of agencies using AI for media planning, buying and optimisation. At least they say they are. I have tested ChatGPT to build a media recommendation; if you use the correct prompts, you can develop a pretty accurate campaign recommendation.
Formulaic approach
Once again, I would not rely solely on the AI to do your plans, since it kicks out a formulaic approach. I recommend you use the output as a sounding board to see if you may have missed anything. It can also be used as a brainstormer to uncover new channels you might have missed.
But it’s still the role of a human to digest and aggregate these ideas together in a more creative fashion. You should still add your value as a media expert on top of what the machine pulls together for you.
Media optimisation is an area where AI would be super-valuable, but the requirement is to have the output re-inserted into the system where your media is running. For AI to be effective in optimisation, it has to be offered at the source of the inventory. An external AI cannot create the data feedback loop required for the campaign to optimise itself, and I fear that in a fractured marketing and media tech stack, AI may not be the perfect solution.
The revolution may be here, but I think it may still take some time to permeate the business in the way that we hope.
This story was first published by MediaPost.com and is republished with the author’s permission.
Cory Treffiletti is SVP at FIS. He has been a thought leader, executive and business driver in the digital media landscape since 1994. In addition to authoring a weekly column on digital media, advertising and marketing since 2000 for MediaPost‘s Online Spin, Treffiletti has been a successful executive, media expert and/or founding team member for a number of companies, and published a book, Internet Ad Pioneers, in 2012.
