Most marketers by now are at least incorporating ChatGPT on some level, ranging from writing actual copy to using it as a tool for brainstorming and planning. In this, we’re going to look at how LLMs like this can be a helpful tool and, equally as importantly, how they can lead you very awry. 
 
Our point of view? Just like the invention of Excel didn’t replace accountants but instead made it a helpful tool for them to more efficiently do their job, certain usage of LLMs can be similar for marketing your business — just as long as you use your critical thinking along the way. 

AI availability in the Meta UI (and other platforms) 

In a recent blog of ours we looked at some of the features Meta’s offering in their current platform, including machine learning applied to audience targeting, placements, and even copy variations through Advantage+. 

Since publishing that, Meta has fully launched a generative AI for creative assets called Advantage+ Creative, which allows for enhancements — you can toggle these on or off depending on your preference. Notably, these include expanding an image, adding dynamic overlays, and image animation, which in theory can be used to alter new creatives without having to wait on design mockups. 

At the same time, it’s worth acknowledging that it’s not super advanced currently — implicitly, understanding these limitations means understanding an overreliance on these generative features probably aren’t in your best interest. 

Some advantages but you are taking your hands off the controls.  

Helpful ways to apply certain AI tools to your digital marketing 

Copy and content help: We’ll start here because it’s the obvious one — many marketers are already using AI to help them with content. Tools like ChatGPT and Claude can help kick start things when you’re stuck with product descriptions or ad copy, for example. From there you can use that as inspiration and tweak the tone, length, style, and anything else with the messaging or create multiple versions for A/B testing. 

Try this prompt: “Create 5 variations of ad copy for a Facebook ad promoting [Product/Service]. The goal is to increase click-through rates. Include a hook, emotional appeal, and a clear call-to-action. Target audience: [describe persona briefly]. Keep each under 40 words.” 

Audience targeting ideas: . AI can’t guarantee the interests or demographics it suggests will match what’s available in Meta’s Audience Manager, but it can help you come up with new audience segments to test. 

Try this prompt: “Suggest 5 audience targeting ideas for a Meta ads campaign promoting [Product/Service]. Include potential interests, behaviors, and demographics. The product solves [insert pain point], and we’re targeting [describe general customer type].” 

Creative concepting: Similar to the copy and content help, this can function as a tool to bounce ideas around, brainstorming concepts, hooks, copy, and more. It can be a tool to support human-powered creative direction. 

Try this prompt: “Brainstorm 3 creative ad concepts for a paid Instagram campaign promoting [Product/Service]. Each concept should include a suggested visual theme, headline, and brief caption. Goal: to drive engagement and conversions from [target audience].” 

Competitor and market research: When used smartly, certain machine learning can help summarize customer personas, reviews, public competitor data, and crawl, for example, YouTube videos to gather a better picture of consumer sentiment. It can also help pick up on more subtle trends using (add something paid social related).  

Try this prompt: “Analyze current Meta ad trends in the [Industry] space. What types of messaging, formats (e.g. carousel, video), and offers are competitors using? Summarize top-performing themes and include examples if available.” (Best used in Perplexity for real-time examples.) 
 
 

What to consider when using AI — and when it might actually be counterproductive 

When it comes to creative assets, people are savvy. 

We’ve all seen those AI-generated images where someone has six fingers or an extra foot—but have you come across the Instagram reels challenging you to spot the AI image and you can’t spot the difference between that and reality? The technology is advancing quickly, and soon it may be nearly impossible to tell the difference. New tools and innovations are emerging every day, but so far, AI-generated creative hasn’t surpassed the polish and style that graphic designers can achieve. That said, models like ChatGPT and others can be great for getting started or for producing lower-stakes creative assets, just maybe not for something as high-profile as your Black Friday sale ad.  

Bottom line 

Really, it’s all simple: AI is not going anywhere, so it’s important for marketers to know how to leverage these tools in a strategic and productive manner. Machine learning programs can help with efficiency and creativity, so long as you think critically about how you’re using it and are discerning about information it gives you. Where we’re at now, using LLMs and AI is primarily useful for brainstorming and being a “first-touch”  so it’s critical for marketers to familiarize yourself with these platforms to make sure you’re keeping up with the tools that are available for this process.