How to use AI for SEO without just writing 'Slop'
Too many AI + SEO how-to-guides are focused on crafting clever prompts which supposedly can write undetectable content for SEO with LLMs.
This week’s newsletter is sponsored by North Star Inbound and Promptwatch.
Check out my post on Google’s new brand filter and reply to this email if you have this filter in your account yet.
Too many AI + SEO how-to-guides are focused on crafting clever prompts which supposedly can write undetectable content for SEO with LLMs. This isn’t necessarily disallowed, but I am not going to weigh in on that. AI might be also able to come up with SEO strategies, but not the cutting-edge ideas you can come up with if you apply creativity. I am not going to discuss that either, because your goal is to be as humanly creative as possible. Instead, let’s talk about all the ways you can be using AI to do your job faster and better.
[Sponsored by Promptwatch]
Do you want to be the brand recommended in ChatGPT Shopping?
OpenAI released ChatGPT shopping in April of this year and just this week they announced the release of Shopping Research.
Prior to Research being released, we at Promptwatch analysed 1,085,000 ChatGPT prompts to understand how and when Shopping was being triggered.
In some cases we found that 20% of the time it was being activated. With Research now released, this number will only increase.
With Promptwatch, your SEO and Content teams will have visibility into:
- When and How shopping is being activated
- Where the citations are coming from
- When your brand or competitors are being mentioned
So you can focus your efforts to get high intent traffic this holiday shopping season.
Go here to get a free study AND start a trial today.”
AI is a fantastic research and executive assistant that can perform many simple tasks more quickly and efficiently than most humans.
Keyword research
Instead of spending hours in a keyword tool, feed a topic to an LLM and ask it to generate categories of search queries. As anyone who has tried searching with AI knows, it is fantastic at understanding intent.
If you’re targeting “things to do in Zurich,” ( I will be in Zurich at the Google Search Live event in 2 weeks, so ping me if you are around), it can help you map whether people want family-friendly activities, nightlife recommendations, free attractions, or day trips. These are just ideas that you still need to verify with other sources, but they cut ideation time in half.
I take the AI-generated list, cross-reference it with Ahrefs or Semrush search volume data, and actual searches to see what comes up in real life with these terms. A single prompt like the one above of things to do in a city can come back with dozens of ideas.
Competitor research
AI tools can also be really good at competitor research. For my entire career, whenever I wanted to assess my own sites vs competitors, I mostly relied on manual searching. I would eyeball the sites to find insights, look at more visible sites than my own, and dig into SEO tools. Obviously, I was likely missing more insights than I found.
Doing this with an AI tool is so much better. Take a landing page or article that seems to outrank something you have created and ask the engine why it might be better than yours. There will be inevitable hallucinations, but the answers will hopefully give you some ideas to dig into.
Once you are done digging into those insights and still don’t have a way to move the needle, you can feel more comfortable that you have tried everything possible. Sometimes a site will stick in the rankings no matter what happens because Google just likes it.
Topic clustering and site architecture
Site hierarchy and content depth can be significant when you are trying to create topical authority. If I am filling out all the features of a tool or a library of how-tos, I can give the outline to an LLM and ask it to structure the content and identify what might be missing from my outline.
I can prompt it to create an outline that includes topics, subtopics, and even sub-subtopics. This goes beyond asking it to write content for me, since this is very nuanced, it will ideally cover so many more ideas.
Once I have the structure, I can again feed it all the ideas and ask it to find the gaps. I find this much more effective than using a content gap tool in any SEO platform, because the LLM is analyzing the logic flows rather than the search results.
Internal linking
This works for internal linking too. AI can analyze your content and suggest which pages should link to each other. For an ecommerce site, that might mean connecting your “Ovens” to “Oven mitts,” or “Outdoor furniture” to “mosquito repellent,” connections that might have been missed if this were done manually.
Titles and meta descriptions
On large sites, there is an element of programmatic content creation for titles and meta descriptions, but the nature of find-and-replace will lead to some comical mistakes. Using AI to generate these titles and descriptions will allow for prompts that keep titles and meta descriptions within specific lengths while minimizing broken text. You can also use prompts to optimize CTR with clickable titles that go beyond what find-and-replace in Excel can do.
Content analysis
I export Search Console data and feed it to AI with prompts like “which pages are declining vs last year?” or something more advanced like “which pages with 'vacation’ in the URL have declines of over 10%?” You can do this in a pivot table, but it’s so much faster with a prompt rather than a sort.
Once you start prompting your data, you will come up with many other advanced ways to ask questions and identify opportunities and anomalies.
Behavior analysis
Feed it your Google Analytics data and ask it pointed queires. For a travel site it might be “What patterns do you see in how people navigate my Japan travel section?” You might find that people reading your Tokyo guide almost always visit your JR Pass guide, which means you should link them more prominently.
Or it might notice that Southeast Asia SIM card guides have high bounce rates, but Europe SIM card guides don’t. That’s a signal to look for new ideas to improve performance.
Other ideas
LLMs could also be great at analyzing competitor XML sitemaps to identify content gaps on your own site, creating PR target lists for citations, and even drafting outreach emails. LLMs can help you find new target markets, draft surveys, and calculate TAM.
Don’t be limited by thinking LLMs are just suitable for writing content for SEO; there’s so much more you can do. The biggest mistake you can make is using AI as a replacement for your brain instead of a tool to help it.
You’re the expert, LLMs are the tool. AI is not going to replace good SEO, but it can help good SEO become great SEO.
Brought to you by North Star Inbound—the sales enablement SEO agency.
Drive High Intent Leads with SEO.
What happens when you combine best-in-class SEO with conversion optimization?BigRentz’s traffic increased by 186% (85k), added 1950 conversions in 12 months.
Self Financial’s traffic increased by 50k/month, and added 685 new customers.
Lastly, Secure Data added 1968 phone calls.
North Star Inbound’s SEO strategies earn leads, conversions, and revenue..
Book a call for a free content audit and 10% off any engagement.
Intrigued? [Learn more here]