How marketers should use ChatGPT and generative AI to build campaigns

ChatGPT and generative AI will transform marketing. Advertisers can use the tech now to help create campaigns, from research to output to iteration. “Nobody should be doing [marketing campaigns] fully manually,” said Mike Kaput, CCO of the Marketing AI Institute. Here are the steps to use AI to build campaigns.

Ideate: One of ChatGPT’s strongest uses right now is brainstorming. “I’d seriously consider it for actually creating ad copy for ad headlines, or at least getting a bunch of different ideas that you can spin off,” said Kaput.

Use ChatGPT to ideate campaign concepts, headlines, copy ideas, or even tone suggestions. Then work with humans on determining the best direction.

Draft: Once you have a campaign idea mapped out, generative AI can bring it to life. “Theoretically, you can use generative AI tools to do basically any type of content writing or advertising copy task you can think of, and it can do it today, I would argue quite well,” said Kaput.

This is a place where Kaput warns of hallucination risk, where ChatGPT will confidently state false information, as well as copyright issues. Content marketers also need to find ways to stand out, because if you can use AI to churn out hundreds of SEO-optimized articles, so can everyone else.

Iterate: Use a combination of copywriting AI and image-generation AI to create different versions of an ad. Adjust mood and tone so you can test multiple ads in different places to see which is the best example.

Test: Platforms like AiAdvertising can use AI to predict ad performance in advance. These sorts of platforms tend to be a higher investment, noted Kaput, so smaller brands may have more luck taking advantage of programmatic tools from Facebook and Google.

Know when to outsource: Marketers can do a lot of ad creation with generative AI themselves, but there are limitations.

  • Plugins: A number of companies are already piloting ChatGPT plugins. Expedia has one for planning trips, Instacart can help suggest recipe ideas and build shopping lists, and Klarna can parse through online shops and compare prices. Kaput sees plugins as the next App Store and a space for brand investment. “That might be the first person I start paying,” he said.
  • Personalization: AI has the potential to unlock hyperpersonal campaigns. “Generative AI, if it’s handled right, will have the ability to figure out who you are, where you are in the stage of the journey, determine what you are interested in, and serve [you] a context” or personalized images and copy, said Dhruv Grewal, PhD, professor of ecommerce and marketing at Babson College. But that level of specificity will require technologists and experts.
  • Integration: Tech experts can help your team get AI products to work together. This is a particularly useful place to invest in people who are willing to experiment, because the tools are changing so fast that what works today might not be as effective tomorrow.

Tools: “ChatGPT is kind of like the shiny object,” said Kaput. But there are a number of other tools worth looking into. Kaput suggested, Jasper, and for text; Grewal suggested Stable Diffusion and Dall-E 2 for image generation. Knowing which tools to use will take testing, but marketers are fully capable of doing this experimentation without being tech experts.

This was originally featured in the eMarketer Daily newsletter. For more marketing insights, statistics, and trends, subscribe here.

"Behind the Numbers" Podcast