Keyword Research in the Age of AI: A 5-Step Strategy

Keyword research [music] is the foundation of SEO. And if you've tried it recently, it feels broken. The keywords that used [music] to work, informational how-tos, and what is queries, have gotten eaten alive by AI. In fact, even if you rank number one, you're losing around 35% of your clicks to AI overviews. That's terrifying. But keyword research isn't dead. [music] It's just evolved. And if you follow my five-part strategy, you'll be ahead of 90% of SEOs who

haven't caught on yet. And it all [music] starts with a 10-second hero prompt. When Chat GPT launched in 2022, people tested it for everything, including keyword research, and the consensus, total trash. It sped out generic content, made up keywords, and invented fake search volumes. But here's what most SEOs won't admit. It wasn't AI's fault. It was the prompts. When you give AI nothing to work with, you get nothing useful back. And that hasn't changed. To find

relevant keywords, you need [music] two things. First, you need seed keywords, which are broad terms related to your niche. And second [music] you need modifiers, which are add-ons like best or howto that turn these seeds into real searches. Combine good seeds with good modifiers in a keyword research tool, [music] and you'll uncover a treasure trove of keywords people actually type into Google. That's exactly what the 10-second hero prompt does. Just open your AI assistant of choice

and say something like, "I'm doing keyword research for my coffee review site, which makes money through ads, sponsorships, and affiliate revenue. My target audience are new and aspiring home baristas. [music] Give me 10 seed keywords that are 1 to two words max, and five plus modifiers that will help me surface appropriate content formats I can use in my keyword research. The seeds and modifiers should not share the same words. And now we have a solid list

of seeds and modifiers, which builds the foundation for everything that comes next. But these are just the starting point. The second part of this strategy is the keyword multiplier. Just take your seeds and paste them into a keyword research tool like HR's keywords explorer. Go to the matching terms report to see tons of keyword ideas. Then add your modifiers using the include filter. [music] And just like that, you should have hundreds, maybe even thousands of real

keyword ideas your audience is typing into [music] Google. But here's the problem. Some of these keywords are going to look incredibly enticing. High volume, high traffic potential, low keyword difficulty scores, all the right metrics. But some of them are a trap. The keywords that used to work, informational how-tos, what is queries, even some best of searches, are getting eaten alive by AI. You can rank number one and still get almost nothing because Google already answered the

question. So, you can't just pick keywords that look good anymore. You have to vet them harder than ever. And you can do that with the bid method. The bid method makes sure that you don't fall into [music] one of these traps. Before you commit to any keyword, it needs to pass three tests. First is the business potential test. Ask yourself, if I rank number one for this keyword, does it actually [music] help my business achieve its

goals? Like, let's say that you have a review site on espresso machines. A keyword like what is espresso has solid volume and very low difficulty. But think about what people searching for this [music] actually want to see. They're not looking to buy anything, maybe ever. They just want a quick answer. Compare that to best espresso machine under $500. Way less volume, but anyone searching that knows they want an espresso machine and they've set a budget. Always

choose keywords that move the needle. Test number two is to vet intent. Listen, no matter what you do, if you can't match the intent of the searcher, you will never rank. To check intent, just Google the keyword and look at [music] what's actually ranking. For example, search for espresso tamper and you'll see every top result is an e-commerce product or category page. This tells us that people searching this want to buy, not read a review. So,

if you're trying to rank a blog post here, not going to happen. The SER gives you insights into what searchers want. match the intent or move on. And the final test is the difficulty check, which just might be the most important of the three. You need to choose keywords you actually have a chance at ranking for. HF's free keyword difficulty checker is a good place to start. [music] Just search for a keyword and you'll see a

score from 0 to 100. The lower the score, the easier it is to rank. But let's be real, there's no way a single number can capture the true difficulty of ranking in an algorithm Google has spent billions perfecting. That's why you need to dig deeper into a page's metrics. You [music] can do that again with the free difficulty checker. First, check the referring domains column. This shows how many websites link to each ranking page. More domains

generally means a stronger page because backlinks are still one of Google's most prominent ranking signals. Second, look at the DR or domain rating of the top ranking pages. This metric represents the strength of a website's backlink profile. Higher DR usually means a more authoritative site, which means tougher competition. In this case, we've got a couple low DRs in the top [music] 10. That's a good sign. Now, these checks aren't foolproof. Remember, billiondoll algorithm versus us. But

the more boxes you tick, the better your odds of finding lower difficulty keywords. By the way, if you have an HFS account, you can filter for all of this at once. Set a max KD of 20, add a lowest DR filter to around the same, and you've got a list of keywords that pass all three tests. So, that's bid, business potential, intent, and difficulty. If a keyword passes all three, you should consider targeting it. Unless the

AI overview is so good, there's no reason for anyone to click through to your site. Remember, Google's job is to give the best result for any query. And there's hard evidence that for a lot of queries, AI overviews are the best result. That's why clicks are down 35% to top ranking pages when an AI overview appears. And it's because of a hard truth most people still won't accept. AI is better than humans at explaining stuff. If

a query has a simple surface level answer, no expertise required, no nuance. There's zero reason for someone to click anywhere. Google already gave them what they needed. Mission accomplished, query resolved, no click required. So before you commit to any keyword, just Google it. Look at what shows up. Put yourself in the searcher shoe and ask, [music] "Am I satisfied with this answer or do I need to click somewhere to learn more?" If the AI overview nails

it, that keyword might be a trap. But here's the good news. Even though AI is eating clicks for a ton offormational keywords, there's still a whole category of searches AI hasn't touched for now. Let me show you something interesting. Back in 2020, we had an evergreen blog post ranking consistently near the top of Google for years. But as chat GBT launched and AI overviews rolled out for more queries, our traffic to that post dropped [music] without

it around 77% from peak to trough despite our rankings staying flat. Now contrast that with this page. Traffic has stayed mostly flat throughout. The difference, this isn't a blog post, it's a free tool. Search backlink checker, no AI overview, mortgage calculator, nothing. Word counter, nothing. Random name generator, nothing. Why? Because when someone searches for a tool, [music] they need to actually use something. AI can't replace that yet. So, here's how you can find opportunities for your

site right now. Go to keywords explorer and start with broad seeds related to your niche like coffee and espresso. Head to the matching terms report and add modifiers like calculator, checker, generator, and more. Look for keywords with decent traffic potential that you could realistically build. Hire a developer or ironically use AI to build it for you. One good tool can outperform 50 blog posts in both traffic and links. So don't sleep on this. Now tool keywords

seem AI proof for now, but there's a bigger shift happening not just in SEO, but across the entire digital marketing landscape, and it all comes down to the fifth [music] and final piece of the puzzle, brand. There's an electric bike company called Electric. A few months [music] ago, they sponsored a YouTube series by Ryan Tran, one of the biggest creators on the platform. But instead of a typical ad [music] read, they donated a bike and set

up a series of challenges throughout the series. And something crazy [music] happened. People started talking, not just about Ryan, but about Electric. These conversations exploded across Reddit, Twitter, YouTube comments everywhere. Now, here's what most people miss. These conversations aren't just word of mouth. their training data. Reddit sign partnerships with both OpenAI and Google. Every post, every comment, every brand mention is feeding the AI systems that power search. And not so coincidentally after the sponsorship, Electric's mentions

in AI overview spiked. Their impressions skyrocketed. And the same happened in Chat GBT. So now when people search best electric bike in Google or ask Chat GBT, what's the best electric bike under $2,000? Guess who keeps showing up? Electric. This is the new game. The more often AI sees your brand connected to a topic, the more confidently it recommends you. Now, this might not seem related to keyword research, but it is. If you flip the model,

instead of asking, "What keywords should I rank for?" ask what queries do I want my brand to be associated with in AI search? Then work backwards. Say you're an electric bike company competing with electric. You'd want to find every query where they're being mentioned in AI overviews in chat GPT so you can get mentioned there too. Go to HS brand radar, enter your brand, your competitors, [music] and your niche. Hover over your brand and click the

number beside others only. And now you have a list of every keyword in prompt where your competitors are showing up in AI and you're not. Keyword research isn't just about finding a topic and matching intent anymore. You need to think about two systems. Google search, which has changed but is still alive and well, and AI search, which is deciding who gets recommended and cited. Master both and you'll be ahead of 90% of people still doing keyword

research like nothing has changed. And if you want to go deeper on ranking and AI, watch our GEO video where I break down the top ranking factors for AI search. I'll see you there.