| Hey good folks and happy Thursday! I hope you’re enjoying this third day of fall! It’s still in the 90s here in Austin, but I swear the breeze before the run rises is getting cooler, so fall (even here!) is around the corner. Let’s dive in.
If you’ve been reading for a while, you know that my team has built a custom AI content production tool in-house. Right now, it sits on top of Claude, but we’re adding in ChatGPT and Gemini here soon. We’ll test different LLM usage at different content production steps (outlining, creation, validation) to increase output fidelity. As a result of this work, I’ve become a bit of a product manager on top of my regular content strategy work. Here are the top questions I ask when working with the larger team on using our new AI tool to automate and scale: Where does the work currently start? Is there a ticket on a project management board? Is it a slack thread? How do you know when you need to create something? And then, where do you document the details on what needs to be created and by when? AI should supplement your existing process, not force you to create a new one. So, knowing where you already start is critical for getting AI to work with you. What information is critical to know in the creation of the asset? This includes the type of content, the format, any relevant product or company information, etc. Where is this information documented? How often is it updated? How often do you need to reference it? This helps us determine if we need to create a skill, or just input a Google doc for additional reference material. Where are the humans in the loop? Most folks are using AI in some way at my current company, so I want to know what they use AI to do, and what they still do on their own. From there, we can figure out where AI can come in to help scale, and where we should keep humans in the loop to ensure high-quality outputs. This will likely need to be revisited as the teams get more adept at the process or as new models come out. My best piece of advice here: humans should be involved heavily at the beginning (briefs and outlines created without AI), at the end (copyediting and general readability edits without the use of AI), and finally in QA (double checking that everything looks good on the page prior to publish). What does the process look like today? This involved breaking down the steps required to create an asset today so that we can understand our way of thinking about that process. With AI, it’s likely we’ll change the entire process, but understanding why we built the process the way we did without AI is a critical step to understanding how we use AI as a tool to scale quality.
AI is a wonderful tool. I genuinely like the challenges it has posed for me and my team, and the way it’s forced a rethinking of how I’ve long created content and content programs. But, AI is only as good as the thinking that goes into it, especially your own thinking on how to piece together something that hits your goals and genuinely serves your audience. Hope this helps!
A NOTE ON THIS ADVICE You do you! One content marketer’s best practices aren’t always right for another one, though I do try to distill out the main concepts and core practices I believe everyone can benefit from. That said, you must use good judgment when deciding whether to take advice given from folks on the internet. I am an expert, and this advice comes from my direct experience, but I am not smarter than you, and I have nothing to gain or lose because of what you do.
Thank you so much for reading. Let me know what you think by replying to this email. Very excited to be here with y’all. Tracey |
