How to Humanize AI Content (Before It Destroys Your SEO)

All right, party people. This is about how to reduce obvious AI writing, especially when you're doing search engine optimization. If a searcher lands on your page and feels that it was written by AI, this decreases trust because a searcher is wondering, are there hallucinations here? Am I going to learn anything unique? AI just parrots the status quo. AI parrots what is popular information, not unique information, and not necessarily the most effective information either. So, it decreases trust and that decrease of trust increases pogo sticking. Pogo sticking is a horrendous SEO signal. You want to avoid pogo sticking. It's when a searcher goes to your page, says, "This isn't for me." Bounces back to the search results, and clicks on another result. That is super bad for SEO. Obvious AI writing decreases conversions. Again, the visitor is wondering if there are hallucinations. So, is the product being recommended in an accurate way or not? I don't know. But it also seems less authentic. People like to buy from other people, especially when you're on a company website or you want your company to have a brand voice. You want it to have a unique voice. And when it's obvious AI writing, it's less likely to have that voice. Now, it's different. You know, people care a lot less when they're using an LLM to get product recommendations because then at least they know that this is an AI. But if you're doing SEO and you have obvious AI writing, it can really hurt your results, hurts your rankings, and hurts your conversions. And so I came across this awesome post from Leila Hermosi, an internal memo, and she said, "I'm going to be direct. I am so sick of reading AI slop, especially in memos." So I would like to explain how you can immediately tell whether a memo is written by AI or a human. At the end, I will tell you what to do instead to write your memos. If we want to be the market leader and help our clients and partners, we need to lead by example. This was for acquisition.com. And if it's good enough for acquisition.com, probably good enough for you. And I'm also making this episode because one of my most viewed podcasts is an episode called how to humanize chatpt content. That was episode 675 of this show, but I made that a year ago and the rules have changed a little bit. AI writing has evolved. So, we're going to start with some of the obvious things to do, then some more recent methods for humanizing AI content. And I'm also going to give my tips on how to humanize AI content because, you know, surprisingly, a lot of people don't know this. I use AI for writing a pretty good deal, but nobody would know because I humanize it very well. So, Leila starts the top five phrases that instantly tell me AI wrote your memo. Number one is delve into slash unpack. And she gives an example. Let's delve into the Q1 results and unpack what this means for our strategy. She says nobody says delve in real life. You'd say look at or break down. Number two, this signals that slash this underscores. And she gives an example. This signals a shift in customer behavior and underscores the need for a revised approach. She comments that's AI connecting two ideas for you because it has no actual opinion. A human would say customers are doing X, so we need to change Y. I'm going to be fair. I actually say this signals that sometimes I probably say underscores a little bit. I do say signals a good deal of the time. I talk about signals with SEO, SEO signals. I already use the word signals in this podcast, but I do say this signals that. The thing the thing to keep in mind with all of this is if you have just like one thing that AI does, but everything else seems normal and human written, it can be okay. But if you have several of them together in one piece of writing, it's going to be a lot more obvious that the writing was done by AI. Again, a little of these things here and there, it's okay. Taken together, pretty obvious AI writing. Number three, navigate the complexities of or in an everchanging landscape. The example is as we navigate the complexities of scaling in an everchanging landscape. And Leila comments this says absolutely nothing. What complexities? What is changing? And she's right. It's not specific, but also it is super super common in AI writing. Number four, synergies or leverage our learnings or holistic approach. By leveraging our learnings, we can unlock synergies across teams and take a more holistic approach. And she writes, "This is corporate baloney. It sounds profound and means nothing, and nobody who reads it has a clue what it says. A human would say, "Here's what worked. Here's what didn't. Here's what we're changing." I have an awesome list of chat GPT buzzwords from Reddit. I'm going to share these. Some of them were already shared by Ila. Just a good list of chat GPT buzzwords to try to avoid. Embarked, delved, invaluable, relentless, groundbreaking, endeavor, enlightened, insights, esteemed, shed light, deep understanding, crucial, delving, elevate, resonate, enhance, expertise, offerings, valuable, leverage, intricate, tapestry, foster, systemic, inherent, tapestry. again. Treasure Trove, Testament, Peril, Landscape, Delve, Pertinent, Synergy, Underscores, Empower, Unleash, Unlock, Elevate, Foster, Delve was using that many times. That list came from the Chat GPT subreddit. Again, if you use one of these words once, not many times. It's okay. But if you're using them a lot, okay, this is going to throw off red flags. This next section is the top five things ChachiPT does that give it away immediately. And this is where things kind of changed since one year ago. You didn't see this as much a year ago. Now you see this a lot more. Number one is the transition words. Every single paragraph opens with moreover, furthermore, that said additionally or importantly. Read your draft out loud. If you wouldn't say it, don't write it. And that is great advice. I advocate just in general, read your writing out loud because it's easier to identify things that sound awkward, things that aren't in your voice. Maybe you were tired when you wrote it, things that aren't clear. It's so much easier to find mistakes when you read your writing out loud. I promise you, it will make a huge difference in the quality of your writing and the amount of mistakes that you make. You will make far less mistakes. Number two, and this is a clear pattern that you see all the time. It's you have bullet points and each bullet point begins with a word that is bolded and then a colon after it. So clarity and clarity is bolded and there's a colon after it and then it says ensure your memo is clear and concise. There's an explanation after each bolded word. Alignment colon and alignment is bolded. Make sure all your stakeholders are on the same page. Execution. Execution is bolded. There's a colon after it. Focus on actionable next steps and then all of these are bulleted. I use bullets all the time, but I do not structure my bullets like this. I wish I could. It is very clear, but it reads like obvious AI and I don't want to reduce trust for me. The tradeoff of appearing authentic is well worth it. Number three thing Chat GPT does that gives it away immediately. The corporate therapist voice. The example is this is a powerful opportunity to lean into our strengths and foster a culture of accountability. Yeah, that is terrible. Number four, the neat little bow at the end. The example ultimately the goal is to build a more resilient and agile organization or at the end of the day it comes down to execution. Here's why this is bad. If your conclusion could apply to literally any company on Earth, it just doesn't really say anything. And number five is the says everything means nothing BS. You read an entire paragraph, it sounds smart, and then you realize you can't summarize what it actually said. That's because AI is pattern matching language, not thinking. If you read your own memo and can't tell someone the point of each paragraph in plain English, delete it. No memo is better than a poorly written memo. And that's an interesting point. You know, you could actually say that with SEO content because if you put out enough bad SEO content, there's such a thing as a sitewide quality score. And if people keep pogo sticking from your SEO pages, Google's just going to stop trusting your site. You're going to have a lot of trouble ranking in the future. And just with any normal visitors, you're going to reduce trust a lot and people might not give you the time of day in the future. So this is what to do instead. Again, this is from Leila Heroszi who suggests use a voice dictation technology to write your first draft. Talk through your idea out loud. Say it in your own words with your own voice. A voice dictation technology transcribes it. That's your raw material and it already sounds like a human because it came from one. Guess how I wrote this? I didn't. I wrote it with my voice and edited it after. So, interestingly, when I read this was in the slide before, there were two things that sounded like AI writing to me. If you wouldn't say it, don't write it. And if your conclusion could apply to literally any company on Earth, it's not a conclusion. It's wasted words. That might have been made by AI. However, to my earlier point, because this whole thing isn't filled with AI giveaways, I trusted it a lot more. I felt like it was original thinking that was cleaned up with AI. Something else words were just capitalized to emphasize things. You never see AI doing that. Lots of words in this are capitalized to emphasize the points that Leila is trying to make. And my point saying all of this is if there are a few little giveaways here and there, it's okay when taken as a whole. If the whole thing largely seems unique, people will trust it a lot more. And Leila even says she she might have used people like Whisper Flow for voice dictation. She might have used something like that. She might have just used her computer to transcribe her voice, just her native computer's functionality. But yeah, a few giveaways here and there. Not a big deal. A ton of them. That's when it gets bad. Number two, for what to do instead, use AI to make it better, not to replace your thinking. Once you have your draft, ask your AI to help you by asking, "What is missing from this memo?" "What questions would a reader still have?" or "Where is my language unclear or ambiguous?" or "Help me tighten this." Make the language more concise without changing my voice or, "Does my argument flow logically?" I actually do those things all the time with my writing. Number three is do not paste in a topic and ask AI to write the memo for you. And this last part comes from Leila's husband, Alex. And Alex Heroszi wants to add the forced negation. Not this, but that. Kind of what I said earlier. And in instead of X, Y, or the staccato repetitions, not this or this or this, but that or excessive adverbs, quietly underscores, just say this is important. All of that is AI slop. And the too long didn't read from Alex is the bullets you sent to AI to write your memo are arguably more valuable than the memo. I'd rather get those than the slop. And that's also a very good point. Sometimes you give the most straightforward, easy to understand explanation, and an AI can actually make it harder to understand, more verbose, more wordy, longer, unnecessarily longer. So, a few honorable mentions for how to humanize your AI content. Remove M dashes. M dashes, those long dashes. Still very, very clear AI writing giveaway. I replace all M dashes with space, normal dash, and then a space. I just find all and then replace. Something else. A lot of people don't know this. Add images and videos specific to your content. Especially images and videos that are not AI generated. So these could be screenshots or screen recordings or even just extremely related YouTube videos. [clears throat] I think part of what it is is that multimedia distracts people from the AI writing. That's a big part. But also when you are not using generic AI generated images and video, it really just feels more natural. And the third thing, this comes from Justin Fineberg on Instagram who shared this chat GBT behavior prompt to humanize your text. So the prompt is and this is a highlevel prompt within personalization and you say write like a human keep it professional but conversational don't use m dashes or buzzwords like streamlined avoid sounding like a press release be clear direct and natural like you are writing to a smart friend and that will reduce a lot of problems you have with AI writing having that in your personalization settings again obvious AI writing decreases trust increases pogo sticking and decreases is conversions. So, if you can avoid these obvious signs, you will get crazy benefits. You know, a lot of people are using AI to automate parts of my SEO method, compact keywords. That's the thing on my shirt. It is my SEO course that lots of people are loving. The reason that they're able to automate parts of it is because the course is so clear. I've gone out of my way to make it so clear and easy to implement. And the thing is, once you fully understand the method, you can use AI to improve your SEO landing pages. But you have to first have that baseline understanding of what to do. It's similar to how giving your thoughts and having AI help you clean up your thoughts. You you verbalize your thoughts. It's transcribed. AI helps you clean it up. With the course, you understand the method. You're able to do it yourself. AI helps you do it better or saves you time in some areas. The course is specifically about getting customers with SEO, getting users, getting warm leads calling you up. And if you haven't checked it out yet, you will love it. It is at compactkeywords.com. I hope you will take a look. And that is all for episode 999 of this podcast. 999 days in a row doing this show. Episode 1,000 coming up tomorrow. No days missed since I started this. I haven't missed a single day of recording since I started this podcast 999 days ago. If you watch this on YouTube, thank you so much for watching. If you listened on Spotify or Apple Podcasts, thank you so much for listening and I will talk to you again tomorrow.