Teams waste time mass-producing AI pages without strategy. Real content needs information gain: unique context search engines haven't seen. Ground your AI in what your audience actually cares about and high-quality data sources, or you're just scaling garbage.
There's a lot of talk about like, hey, now, you know, I could just publish 1,000 pages now and let AI do all the work. So like, where, from everything you're seeing with your customers, where is content engineering, where is that approach like genuinely a step change? Where are teams kind of fooling themselves?
Yeah, great question. It's probably the one I get the most. And it's, it's so interesting that it comes from people who are just dipping their toes in the water. And it comes from really big companies of like, where's the upper limit on like, how much I can let the AI take over. But if we take one step back of like what the intention of creating content is, and I think Eli Schwartz in his book Product-Led SEO makes some really good points and it's very like sentient. It's written 10 years ago before a lot of the different real estate even opened up in Google. But he makes the point of the search engines, their full intent is to surface the best context and content to the user. So if we bring that to today, whether it's ChatGPT or Perplexity or Google and all of their latest features, that, that's still what they're trying to do. They're trying to find for your given question or prompt what's the best, um, piece of content to go surface for it. And what that needs to be grounded in and what we're seeing success from like our customers and our own content and what you can do today is like making sure that you're providing information gain to the LLM. So what does that mean? That means that you're giving it context that like the LLM isn't trained on. Like if you can imagine OpenAI is going to scrape all of the web pages out there every couple of months, like that becomes like what the average marketer or what the average prompt will return. The search engines reward today is like unique content, whether you're on my side of the world, which is like content, or you're like a more of a deep in the weeds GTM engineer on like sales, you wanna make sure that you're bubbling up the right context and that's gonna be totally grounded in like what your audience cares about. So we can go deeper into like what data mining is, where to find really good information, but that is what you want to ground all of your work in. And then you can start to use well-constructed agents on top of it to figure out, great, like here's some questions, they're topically relevant, let's create some pages.
Roll Claude out from one laptop to the whole GTM org | Sam Gong & Zach Vidibor
