Imagine you’re the owner of a local plumbing company. You’ve invested months into your SEO. You’re ranking number one for “how to fix a leaky faucet.” Google Search Console shows 1,400 visits a month to that page. Your agency sends you a report with a green arrow pointing up. Everything is looking good performance-wise, but your phone isn’t ringing.

The problem isn’t your ranking. It’s the keyword you’re focusing on.

Type that query into Google today, and you’ll get a complete AI-generated answer at the top of the page, showing you quick step-by-step instructions, tool recommendations, DIY videos, and common mistakes to avoid. Basically, everything a homeowner needs to handle the faucet problem by themselves.

Now you find yourself in a situation where you own a page that ranks first for a query that no longer requires a plumber.

This is the real SEO problem most small businesses are facing right now. AI has started absorbing entire classes of informational keywords, and many businesses are still investing in them as if nothing has changed.

This is already visible in the data. When AI Overviews appear, organic click-through rates can drop by over 60%, even for top-ranking pages. At the same time, more than 80% of searches now end without a click at all, meaning users get what they need directly on the results page. And for queries with AI summaries, about 1 in 4 searches ends without any further action.

But not every keyword is affected equally. Some have been fully claimed by AI. Some are changing shape. And some still drive calls, bookings, and revenue because they require a real business to finish the job.

This article gives you a framework to tell them apart: a three-tier system for classifying your keywords based on how much AI has claimed them, a step-by-step audit to find your most valuable SEO keywords right now, and concrete examples from five SMB industries showing exactly which keywords are worth your time.

Contents

• Why you need to evolve your SEO keyword strategy
• The 3-tier framework to guide your keyword strategy now
• How to find SEO keywords for the AI search era (step-by-step)
• What this looks like in practice: Updated keyword strategy examples
• What to do once you’ve found keywords worth targeting
• A simple pressure test for any keyword

Why you need to evolve your SEO keyword strategy

Most small businesses experience traffic drops and assume the same thing: something went wrong. Maybe rankings slipped. Maybe Google changed the algorithm. Maybe a competitor overtook them.

Sometimes that’s true. But more often now, it’s nothing from above. The keyword just stopped behaving the way it used to. That’s the part that’s easy to miss.

But the role that the keyword plays in the journey has quietly changed. And unless you separate those changes, it all gets lumped together as “traffic loss,” which leads to the wrong fixes.

Some queries have been fully claimed by AI

There’s a category of keywords that no longer really belong to websites. These are classic informational queries: “what is,” “how to,” “why does.”

A few years ago, these were the backbone of most content strategies. You’d write a clear explanation, rank for it, and bring people into your site early in their research. Now, when someone searches “what is a dental crown” or “how to fertilize your lawn,” they often get a complete answer before they even see a list of results. Structured, readable, detailed enough that most people don’t need anything else.

So even if your content is technically winning that keyword, it’s no longer controlling the outcome. You’re part of the source material, not the destination. Traffic loss in this category isn’t something you fix with better optimization. It’s a structural change in how the query gets answered.

Some queries are partially affected

Not everything collapses that cleanly. There’s another layer of keywords where AI does a decent job summarizing the space, but doesn’t fully resolve the decision. These are the comparison and research queries: “best CRM for small business,” “dental crown vs. veneer.”

You’ll still see an AI-generated summary. It might list a few options and outline pros and cons. But it doesn’t settle the question.

AI Overviews are already reducing click-through rates for many of these queries, but they haven’t killed them entirely. It doesn’t know your budget, your constraints, or your specific situation. So people still click. They open a few tabs. They look for examples. They try to validate what they just read. Fewer of them may make the jump to your website than before, but the ones who do are much closer to a decision. That’s why these keywords still matter.

Some queries still require a business

Then there’s the third category, and this is the one most businesses should be paying the most attention to.

Search something like “emergency plumber Austin” or “dental crown cost Chicago,” and the intent is immediately different. The user isn’t trying to understand a concept. They’re trying to solve a problem. AI can provide context, explain what to expect, and outline typical costs. But it can’t actually complete the job.

At some point, the user needs to call someone, book an appointment, or show up somewhere in person. These keywords still behave much closer to how search used to work. Clicks still matter. Visibility still leads somewhere. Ranking still has a direct connection to outcomes.

Why this distinction matters

If you treat all keywords the same, this change is almost impossible to diagnose. A drop in traffic from an informational query looks identical in a report to a drop from a transactional one. But the underlying cause and what you should do about it are completely different.

One category is gone for good. One is evolving. One is still fully in play. And until you separate those three, it’s easy to keep investing in areas where the return has quietly disappeared while underinvesting in the ones that still drive real business.

🔎 The easiest way to find keywords? Try our Free Keyword Research Tool!

The 3-tier framework to guide your keyword strategy now

Here’s the simplest way to think about your keyword strategy now:

Tier	Keyword Type	AI Impact	SMB Value
Tier 1	Informational/definitional	Fully resolved by AI	Low
Tier 2	Research/comparison	Clicks reduced, but present	Medium
Tier 3	Local/transactional/action	Requires a business	High

The mistake most businesses make is overinvesting in Tier 1. That made sense before. It built traffic. It fed the funnel.

Now, it often produces visibility without outcomes.

The change is simple:

• Reduce effort on Tier 1
• Refocus on Tier 2 and Tier 3
• Judge keywords based on business impact, not traffic

How to find SEO keywords for the AI search era (step-by-step)

Most keyword audits still start from the same assumption: improve rankings, improve results. That used to be true. Right now, you can rank first, bring in consistent traffic, and still get nothing from it.

So the point of this audit isn’t to find where you rank or which keywords you’re missing. It’s to answer one uncomfortable question:

Does this keyword still require someone to contact a business, or can it be resolved entirely on the search results page?

Step 1: Pull your existing keywords (start with what already gets traffic)

Before you open Ahrefs or Semrush, go into Google Search Console and look at what’s actually driving traffic. Not the pages you think should work, but the ones that are pulling in clicks right now.

This is usually where the disconnect starts to show up. You’ll find pages that look like clear wins in your reports: they rank well, they bring in traffic, they’ve probably been highlighted in monthly updates. But when you think about what they actually contribute to the business, things feel off. The phone isn’t ringing. Nothing downstream reflects the traffic you’re seeing.

Most teams interpret that as a conversion problem. More often, it’s a keyword problem hiding behind good-looking numbers.

Step 2: Check what Google actually shows (not what the tool says)

Once you have your keyword list, search each one manually and look at the results as if you were the customer. Not through a tool. Not through a report. Just a clean search.

Take “how to fix a leaky faucet.” A few years ago, that query would give you a list of articles. Now, in many cases, it gives you the answer immediately: a structured explanation, steps, a video. All before you see a single organic result. At that point, your ranking barely matters. The user has already solved their problem.

Ask yourself one question while you’re looking at the results: if I were actually trying to solve this problem, would I need to click anything here? 

If the answer is no, that keyword has effectively been claimed.

💡 Get updated tips for your SEO strategy with our free guide >> How to Do SEO Right—Right Now!

Step 3: Classify the keyword (This is the core of the audit)

As you work through your list, you’ll notice that keywords in the same general topic can behave very differently from each other.

Take the dental example:

• “What is a dental crown” gets a clean, structured explanation that covers definition, materials, and process. There’s no urgency to go further.
• “Dental crown vs. veneer” still shows a summary, but it’s not definitive. You’re comparing options, weighing tradeoffs. That creates space for clicks.
• “Dental crown cost Chicago” is different again. You’re not learning anymore. You’re trying to figure out what this is going to cost you, locally, and who to call.

All three sit close together topically. All three behave completely differently.

Step 4: Pressure-test the keyword (this is where most people get it wrong)

Don’t rush the classification step. A lot of keywords don’t fall neatly into one tier, and small wording differences can change everything.

For example, “dental crown cost” might still nudge someone to book an appointment, especially if the results are vague or location-dependent.

But “average dental crown cost” is much easier to satisfy with a single price range on the results page. If you group those together, you end up making the wrong call about both.

Look at the actual results and think about what the user is trying to do, not just what the keyword says on the surface.

Step 5: Quantify the gap between traffic and value

Once you’ve classified your keywords, compare what the data shows against what the business is actually getting. This is usually where the biggest gap appears.

Keyword	Monthly Visits	Leads	Tier
how to fix a leaky faucet	1,200	0	Tier 1
emergency plumber [city]	90	14	Tier 3

The keyword bringing in 1,200 visits a month is generating zero leads. The one with 90 visits is generating 14. That’s not a conversion problem. That’s a keyword strategy problem.

Step 6: Identify where your time is going

Zoom out and look at your content as a whole. What you’ll usually find is that a large portion of your site is built around informational queries, because that’s what SEO rewarded for years. Those pages were easier to rank, easier to scale, and they drove visible traffic growth.

The question isn’t “how do we improve these pages?” It’s “what role do these pages actually play now?”

Some still matter. They support other pages, establish expertise, and may get cited in AI-generated answers. But they’re no longer where your growth comes from.

Step 7: Decide what to do with each page

By this point, the next steps tend to become clear without needing a rigid checklist.

• Some content you keep as-is, because it still serves a purpose even if it’s no longer a direct driver of leads.
• Some content you improve, especially Tier 2 queries where better positioning can still influence an undecided buyer.
• Some content you stop investing in entirely. Not because it’s bad, but because the return is no longer there.

You’re not trying to win every keyword anymore. You’re trying to focus on the ones that still require you.

What this looks like in practice: Updated keyword strategy examples

Across almost every small business category, the same thing is happening: there’s a layer of content that used to attract traffic and signal authority but now gets absorbed before it can do anything useful. Right next to it, there’s a layer of keywords that still drives real business but gets a fraction of the attention.

Plumbing company

A plumbing business will almost always have content around basic fixes. Things like “how to fix a leaky faucet” or “why is my sink dripping.” These pages used to make sense. They brought in traffic, they positioned the business as helpful, and occasionally they led to calls.

Now, in most cases, they get resolved before the user ever needs a plumber. The answer is right there. The steps are clear. The problem feels manageable. So the business ends up ranking for a query that actively removes the need for its service.

Compare that to queries like:

• “emergency plumber [city]”
• “burst pipe repair cost [city]”
• “licensed plumber near me same day”

These don’t get resolved in the search results. They lead to action. The user isn’t trying to learn anymore. They’re trying to fix a problem they can’t handle themselves.

That’s the difference.

🛠️ Get everything you need to know about showing up in AI search in our free guide: The AI Search Toolkit >>

Real estate

Real estate sites are full of educational content.

• “How to buy a house.”
• “What is a mortgage?”
• “Steps to selling your home.”

All of these topics can now be summarized cleanly in a few paragraphs. AI lays out the process, explains financing, and answers the basic questions most buyers have early on. That doesn’t mean the content is useless. It just means it no longer drives the outcome.

Now look at queries like:

• “homes for sale in [city]”
• “real estate agent near me”
• “how much is my home worth [city]”

No AI Overview for this query (yet)!

These are not informational in the same way. They’re situational. They depend on location, timing, and context. At some point, the user needs a person to move forward.

That’s where the click still matters.

HVAC business

HVAC companies often invest heavily in educational content around systems.

• “How does central air conditioning work?”
• “types of HVAC systems.”
• “what size AC unit do I need?”

These are all logical topics. They help explain the service and build credibility. But they’re also exactly the kind of queries AI handles well. The answers are structured, repeatable, and easy to summarize.

Now contrast that with:

• “AC installation cost [city]”
• “emergency HVAC repair near me”
• “HVAC company [city] reviews”

Here, the user isn’t trying to understand the system. They’re trying to solve a problem or make a decision. And that decision can’t be completed without involving a business.

Dental practice

Dental sites follow a similar pattern.

• “What is a dental crown?”
• “How long does a root canal take?”
• “are veneers better than crowns?”

These topics still have value, but they’re increasingly being answered directly in the search results. The user gets a clear explanation without needing to visit a site.

But when the query shifts slightly:

• “dental crown cost [city]”
• “same day crown dentist near me”
• “best dentist for veneers [city]”

Now the stakes are different. The user is no longer just learning. They’re evaluating options and preparing to act. That’s where the opportunity is.

Car dealership

Dealership content often leans into financing education.

• “How car financing works.”
• “What credit score do you need to buy a car?”
• “lease vs. buy.”

All of these are now easy for AI to summarize. The process, the terms, and the tradeoffs can all be explained in a few lines. But that doesn’t replace the moment where someone actually needs a car.

That shows up in queries like:

• “used cars under $15k near me”
• “car dealership bad credit financing [city]”
• “best used car dealerships [city]”

These queries don’t end in understanding. They end in action.

What to do once you’ve found the right keywords

This is the part most people get wrong. They see traffic dropping on research queries and assume those keywords aren’t worth targeting anymore. So they either double down on volume or abandon the space entirely. Both are mistakes.

AI now handles the first layer of the query: summarizing options, explaining the basics, giving the user a starting point. So if your content is doing the exact same thing–listing options, repeating definitions, staying neutral–there’s no reason to click. AI already did that job.

It’s not just about finding the right keywords; it’s about using them in a way that adds value and earns the click.

To do that, you need to:

• Narrow the problem
• Add real-world constraints
• Show actual numbers where possible
• Make decisions easier, not just more informed

In other words, your content needs to do the part AI stops short of doing because that’s the only part left that’s worth clicking for.

A simple pressure test for any keyword

If you strip everything else away, it comes down to one filter: If the query can be fully answered without a business, it’s not a growth keyword.

That sounds obvious when you read it. It’s much harder to apply in practice, because a lot of content still feels valuable even when it doesn’t lead anywhere. It ranks. It brings traffic. It shows up in reports. But if someone can search that query, get what they need, and move on without ever contacting a business, it’s not contributing to growth in the way most teams expect.

That doesn’t mean you remove it. It just means you stop treating it as something that will drive results on its own.

Before you commit to a keyword, run it through two questions:

1. If someone searches this, will they eventually need a provider to complete what they’re trying to do?
2. Is there any realistic path where they don’t need to contact a business at all?

If the second answer is yes, that keyword has likely already been absorbed. It doesn’t make it useless. It just means it shouldn’t be at the center of your strategy.

Build the right keyword strategy for the AI era

What’s happening to SEO right now isn’t random, and it isn’t temporary. It’s a structural change in how different types of queries get handled.

Some keywords have effectively been removed from the click economy. They still exist, and they still get searched, but they no longer create the same path to a website. Some are still in play, but they require different content to earn attention. And some haven’t changed at all, because they still depend on a real business to complete the job.

The problem is that most strategies treat all three the same, measuring success through rankings and traffic without asking what those numbers actually lead to. That’s why so many businesses feel like SEO is working on paper and failing in reality.

The fix isn’t more content or better optimization. It’s being more selective. Fewer keywords, better aligned with actual outcomes.