In the first version of this piece just two years ago, we wrote that AI marked "the dawn of a new era in content marketing."
If that was the dawn, we're well into the morning now. The coffee's kicked in. Rush hour's in full swing. The kids are at school. The metaphors are belabored...
We've spent those two years experimenting with AI in our own content operations. This guide compiles everything we’ve learned: frameworks for when (and when not) to use AI, strategies for adapting to AI Overviews eating your search traffic, and honest accounts of what we’ve done well and what we haven’t.
Use it to find your balance; where AI saves you time without killing your creativity.
What to Do About Google’s AI Overviews Eating Your Search Traffic
This is the top question for content teams right now. Google's AI Overviews (AIO) are quietly killing click-through rates — down by 15% to 35% in some studies. Odds are, your Search Console is telling the same story.
It doesn't mean search is dead though; it's just different. You need to shift priorities: Put conversion efficiency over raw traffic numbers, cover every BOFU keyword your competitors are ignoring, and build content that turns fewer visitors into more pipeline.
Read AI Overviews Are Eating Your Search Traffic
How AEO Differs From SEO
As you adapt your strategy for the new search era, you'll be dealing with AEO (a.k.a. GEO, a.k.a. LLMO).
Is it the new SEO? Is it just SEO with a different name? Both takes are out there.
Sure, there’s plenty of overlap: AEO builds on the foundation of SEO, and it relies on the same infrastructure, quality standards, and trust signals.
But AI systems process content differently. These models don’t rank pages like search engines. They read, extract, and verify information across sources. We get into how this translates into AEO-specific requirements for content creation, management, and distribution.
Read SEO vs. AEO: A Field Guide for B2B SaaS Content Marketers
How to Build Visibility in AI Search
Now, let’s break down the layers of AEO itself.
We’ve built a framework called the AI Visibility Pyramid that maps the three layers required for visibility in AI search. Each layer builds on the last, and you work all three in monthly cycles: one SEO improvement, one citation-worthy asset, one credibility play.
Read AI Visibility Pyramid: How to Improve Your Presence in AI Search
What We Learned From Three Years of AI Experiments
We're a quality-obsessed content agency, so yeah, you're probably wondering how we square that with AI.
Let’s walk you through our entire AI journey — year by year, experiment by experiment.
We tried an efficiency play that bombed because clients still expected our usual standards. We tried a free-for-all approach that created dozens of inconsistent workflows. Now we're doing something that's working, and we put someone obsessed with editorial quality (not coding) in charge of it all.
Read Our AI Journey: Lessons from Failed Experiments and Where We Are Now
How Content Leaders Use AI
AI has taken root in content workflows, but to make it work without losing quality, you need boundaries and best practices.
We asked five content leaders from Preply, Freed, Semrush, Planable, and KNIME to share their approach to AI workflows — where they use it, and where they don't.
The through-line across all our findings is that you need a strategy for using AI. Consider this your blueprint for building one.
Read 5 Content Leaders Share 7 AI Insights for Your Team
Why You Should Not Outsource Your Thinking to AI
AI promises to cure writer's block. Just press the button and boom: instant first draft.
Don’t do it.
Letting AI fill your blank page is like sending a robot to the gym for you. Your mental muscles atrophy. You anchor to mediocre ideas. You skip the serendipity where your best thinking happens.
Wrestle with the blank page yourself first before bringing AI on board.
Read Stay Strong: Never Let AI Fill Your Blank Page
What It Takes to Succeed With AI Content Workflows In-House
When getting on board with AI, most teams don't realize they're building a mini-product that needs QA, documentation, and someone to maintain it.
AI helps increase output, but it doesn't necessarily reduce work. It saves you time on writing, but then you spend it building workflows, reviewing robotic drafts, and managing expectations.
You need complete workflows (not just ChatGPT in a tab), skilled editors who can tell if content is actually worth publishing, and strategic planning to decide what's worth creating in the first place. Here’s what in-house AI actually requires.
Read AI Content Works (But Only If You Do the Work)
How to Match AI Models to Your To-Do List
You wouldn't hire a single person to be your developer, designer, writer, and analyst. Similarly, using just one AI model is like having a team of one.
Different models give you drastically different results. If you default to one AI model for every task, the TIP Method offers a smarter way. Break down your work into tasks, match them to the level of intelligence you need, and choose based on personality. It will help you find the best fit for each task.
We also touch on why Claude handles single meeting summaries better than ChatGPT Pro, each model’s superpower, and how to avoid switching between AI models too often.
Read The TIP Method: Choose the Right AI Model for Every Task
Finding Your Optimal Human-to-AI Ratio
Use AI as a brainstorming partner and fine-tune the outputs? Feed it your POV and let it handle the rest? Or go against the flow and become a writing purist?
We tried it all, and our answer is none of these. Blending human creativity with AI power isn’t one-size-fits-all. We share how to find the right balance.
Read The Content Cyborg: How to Use AI Writing Tools in Content Marketing
How We Built an SEO Tool With AI
Show us a marketer who hasn't tried vibe coding at this point. Our own Tim Metz built an SEO forecasting tool in Claude's interface in under an hour. It looked great. It didn't work.
He had to move to Cursor to build something fully functional, and it took weeks. But it's still miraculous to build a tool as a non-developer, and the journey was insightful.
Tim covers the full process. He explains why you need at least three different AI tools to pull it off, the challenges you’ll face as a non-coder, and how to avoid headaches along the way.
Read Code Is Now Content: How We Built Our SEO Calculator With AI
Is ChatGPT Pro Worth $200 a Month?
Two hundred dollars for an AI subscription. What would possibly make it worth as much?
We've been testing ChatGPT Pro since launch, and here's what swayed us: o3's deep research rivals a junior analyst, it gives you the ability to process 150-page documents without losing the thread, and using it to vibecode actually works.
Of course, it’s not all good, and we get into the cons, too.
Read The $200 AI Question: Should You Upgrade to ChatGPT Pro?
The Six Prophecies That Explained Today's Content Marketing
Ryan Law wrote predictions for AI in July 2022, four months before ChatGPT launched. He laid out how AI would reshape content marketing from the work we do to the roles we take.
Every single prediction came true. The article perfectly explains what's happening in content marketing right now. Lisan al-Gaib.
Read 6 Predictions About AI in Content Marketing
Vows of an AI Addict
Some days, our team swears by AI and forgets how to write a simple email without its assistance. Other days, we make vows to use it responsibly.
Tim Metz captures that struggle, tracing his own spiral into dependence and how he clawed his way back.
Read Confessions of an AI Addict
Conversations With People Building With AI
We spent a season of the Animalz podcast with leaders who are actually building with AI, not just talking about it.
Some highlights? Nathan Baschez (Lex) on reimagining writing with AI. Kyle Coleman (Copy.ai) on giving marketers their weekends back. Alex Halliday (AirOps) on what he learned from conversations with Sam Altman.
You can find each episode below.
Listen to Introducing the AI & Content Season: Real Talk, No Hype
What Happens Next
AI in content marketing isn't slowing down. Models keep improving, new tools launch weekly, and the gap between early adopters and everyone else grows wider.
We built this resource because we're still figuring it out ourselves. Some workflows we've documented here will be outdated in six months. New risks will emerge. Better practices will replace what works today.
We'll keep updating this guide as we learn. We'll add new experiments, the frameworks that hold up under pressure, and honest assessments of what stops working. Bookmark it, revisit it, and watch it evolve alongside your own AI strategy.