Skip to content

Generative AI Do’s and Don’ts for Better Content Marketing

Technological disruption is par for the course in content marketing. But generative AI hits differently and closer to home.

These versatile tools can speed up tedious production tasks and spare writers the frustration of staring at a blank page. But you can’t just plug them in and sit back while they do all the work.

The value of AI tools depends on the effort you put into them. You can make smart decisions on how to apply these technologies and carefully edit its output to ensure it meets the brand’s standards.  

The value of #AI tools depends on the effort you put into them, says @joderama via @CMIContent. #CMWorld Click To Tweet

What’s the best way to put AI’s content capabilities in play? We asked some experts presenting at Content Marketing World for their top AI tips and recommendations. Here’s what they say you should – and shouldn’t – do:

Do: Use AI for writing and editing assistance

Generative AI tools can help you overcome lank-page syndrome” by acting as your research and brainstorming buddy. They can also help you stay on-brand, providing editing tips for tone and style. Some generative AI tools also provide an SEO assist during the writing process or use your proprietary content to create derivate pieces. – Wendy Covey, co-founder and CEO, TREW Marketing 

The best way to use AI for content can be summed up by this phrase: “Say it better.” There are times when what you’ve expressed, especially in the written word, isn’t what you want it to be – something is missing, confusing, or simply “off.” It’s in these moments that AI has incredible value. –  Marcus Sheridan, vice president, Marcus Sheridan  

Use AI as a way to check that all long-format copy is consistently on brand voice and tone. It’s also a free and easy way to proofread. – Jennifer Harmon, content strategist and creator, Convince & Convert 

#AI can check #content to ensure it maintains brand voice and tone, says Jennifer Harmon via @joderama @CMIContent. #CMWorld Click To Tweet

Don’t: Replace skilled (human) team members

Generative AI should make doing your work easier, not replace the work done by professional creatives. It’s useful for filling in bandwidth or resource gaps – like creating an image quickly instead of looking through stock images for hours. It can suggest titles for articles and subject lines for newsletters if that’s not your strength. You can use it to generate ideas for a topic’s scope to ensure you’re not leaving out an important subtopic. Ruth Carter, evil genius, Geek Law Firm 

Teams that see generative AI as a way to cut staff and increase content volume without making process improvements are doomed. There are reasons why journalists and content strategists have developed processes like style guides, content briefs, fact-checking, and developmental edits. Generative AI doesn’t make those things skippable. – Jeff Coyle, co-founder and chief strategy officer, MarketMuse, Inc. 

Do: Sharpen your audience understanding

Create an AI-powered marketing persona. Check it against your persona research. Improve it until you feel it’s mostly accurate, and then ask it lots of questions:

  • What does your audience worry about?
  • Where do they get their information?
  • What topics are most helpful for their jobs to be done?
  • What triggers them to go looking for answers or help?
  • What information do they need before they’re willing to become a lead?

Unlike humans, you can quiz your new AI persona day or night. It never gets tired of sharing insights with you. – Andy Crestodina, co-founder and CMO, Orbit Media Studios 

Generative AI tends to give good insights into how the audience does research and how to prioritize their needs. For instance, ask ChatGPT, “What are the top features that [our core audience] should prioritize when shopping for [our products/services]?” and you’ll see exactly what points your content needs to address. – Zontee Hou, director of strategy, Convince & Convert 

Don’t: Overlook AI’s legal, ethical, or quality implications

Bias and ethical concerns are associated with generative AI, and there are legitimate concerns over quality, authenticity, security, and privacy. Human oversight and review are critical, and that review requires diverse perspectives to ensure inclusivity and to provide context that AI can miss. By being mindful of these pitfalls, organizations can harness the benefits of generative AI while ensuring responsible and effective implementation. – Karen McFarlane, chief marketing officer, LetterShop x KMC

I’m particularly concerned about the legalities associated with AI on the copyright side and the privacy concerns. If you feed proprietary information into these chatbots, it’s out there for anyone using the chatbot to pull from. Also, on the originality side, if your content sounds like everyone else’s (because all the chatbots can do is regurgitate content), how will that help you stand out? – Michelle Garrett, consultant and writer, Garrett Public Relations 

When you feed proprietary content into an #AI chatbot, the tool can pull from it to answer other users, says @prisus via @joderama @CMIContent. #CMWorld Click To Tweet

Do: Create production efficiencies and iterative assets

I’ve turned our long-form content into short-form videos for YouTube, snippet videos for Instagram, tweets, Facebook and LinkedIn posts, and more. Additionally, using AI to create transcripts for accessibility has been a tremendous timesaver. AI helps me promote our content in more ways and in less time than I thought possible. – Cathy McPhillips, chief growth officer, Marketing Artificial Intelligence Institute 

As a podcaster, I can upload an audio file, and generative AI tools will produce a range of content, including suggested episode titles, timed transcription, show notes, newsletter content, LinkedIn post, tweets, and more. This content still needs some editing, but it makes my job much easier. – Bernie Borges, vice president, content marketing, iQor 

Generative AI is great for content busy work like paid social, email subject lines, tweets, etc. AI can provide various options with the right inputs, but it takes an expert to identify the best option. AI provides a lot of ingredients, but it is not a meal. – Kristyn Wilson, executive vice president of digital PR and communication, Adept 

Don’t: Publish raw AI output

Generative AI is great at solving for the blank page or repurposing one form of content to another as a draft. We’ve also found creative ways to use it to analyze keywords and competitors, outline articles, and create webinar emails and social copy. But copy from AI is starting to look obvious and dull. Avoid publishing AI content without a solid editor juicing it up and adding opinion, humor, and personality. Michael Brenner, CEO, Marketing Insider Group 

Avoid publishing #AI content without a solid editor juicing it up and adding opinion, humor, and personality, says @BrennerMichael via @Joderama @CMIContent. #CMWorld Click To Tweet

Do: Amplify content promotion

One use case I’ve put into practice is writing the social media post copy to accompany an article I’m sharing. I copy and paste the article body to ChatGPT and tell it how many sentences to generate. I do some editing of the provided copy and fact-check it to make sure it doesn’t introduce any inaccuracies. – Dennis Shiao, founder, Attention Retention 

Don’t: Assume AI automatically improves outcomes

Make sure AI is actually improving your processes and output. Track the time it takes you to perform a task and measure it against how long it takes when using AI. – Brian Piper, director of content strategy and assessment, University of Rochester 

Relying too heavily on [AI tools] for writing purposes without putting in the effort for extensive prompt engineering can lead to lackluster outcomes. Large learning models rely on word probabilities, which means that if your prompts lack sufficient detail, you’ll end up with generic results that fail to impress. It’s all about understanding the nature of these models and the importance of providing precise instructions to yield the desired output. – Christopher Penn, chief data scientist, Trust Insights

Relying too heavily on [#AI tools] for writing without putting in the effort for prompt engineering can lead to lackluster outcomes, says @CSPenn via @joderama @CMIContent. #CMWorld Click To Tweet

AI’s potential is in your (human) hands

Take a critical look at your content program to see where you need the most help, then plug AI in to lighten the load. Like any other content marketing tool, the more thought you put into your usage plans, the better the outcomes you can achieve.

Register to attend Content Marketing World in San Diego. Use the code BLOG100 to save $100. Can't attend in person this year? Check out the Digital Pass for access to on-demand session recordings from the live event through the end of the year.

MORE ADVICE FROM CMWORLD 2023 SPEAKERS: 

Cover image by Joseph Kalinowski/Content Marketing Institute