Either fuel for laziness or a tool for expansion. Leveraging generative AI in marketing can’t be avoided if agencies and departments wish to stay relevant in the coming years.
When I first heard about ChatGPT almost a year ago, I was so excited to get my hands on it. My team adopted it quickly and started testing use cases for it. Like many others, my team saw immediate benefits and some very valid drawbacks. So much so, earlier this year we built some material around our use cases so our leadership could go into our community and help educate businesses on how to best leverage it.
After a few months and numerous conversations with my team members, I came to the conclusion that the use of generative AI would become a competitive advantage for advertising agencies and even internal marketing departments. Not to replace humans but to amplify the output of what humans can do.
One specific test that revealed more of generative AI’s potential was observed in our content curation and ideation. We started by coming up with organic social and blog ideas for a handful of our clients. We wrote everything down and documented it carefully including a full social calendar and topic ideas we could research further. Then we asked ChatGPT the same questions we asked ourselves and it outputted roughly 70% of the same information at about 10x the rate. With additional prompts, we were able to refine its output and get a very similar starting point for organic content as we had generated as a team (though research and further editing were obviously necessary).
Another test we ran was within our web development team. We had a specific request from a client on their e-commerce site which we were struggling to execute due to a conflict with the chosen WordPress theme. After working a few hours within ChatGPT, we found its output wasn’t able to accomplish the functionality necessary to satisfy the client’s request and one of our developers wrote the solution within the theme’s PHP files himself. In other more basic situations though, we were able to get custom HTML and CSS written that accomplished the request made by our clients.
We’ve done additional testing across our other teams as well and have made a few conclusions on the current state of generative AI in marketing.
- Treat the output as a first draft, never a final product
- Verify the quality of the output
- Never sacrifice quality for speed
And finally, don’t let stubbornness get in the way of having generative AI help you. I’ve met some resistance against the use of generative AI. Those who may consider its outputs as “cheating”. This mindset I disagree with this as I view it as the next iteration of our ever-evolving industry.
I’m excited to continue testing and using generative AI for our clients as it can help us produce better quality work in less time.