An effective keyword strategy ensures you show up for the terms that matter for your business. Creating one used to be all about finding what your audience searches for and looking at your competitors’ rankings.
But search behaviors and how platforms display results have changed, meaning keyword strategies have to change as well.
Most keyword strategies aren't built for modern search
Keyword strategies need to go beyond traditional Google Search and the typical methods of keyword research to work in modern search environments.
How keyword strategies have always worked
The standard approach to building a keyword strategy typically involves looking at three main areas: competitor rankings, current search results, and keyword databases.
With a competitor analysis, you find keywords your competitors rank for, identify the gaps, and plan to create content to fill those gaps.
Then you check the search engine results page (SERP) to look at autocomplete suggestions, People Also Ask (PAA) results, and related searches to expand your list.
Finally, you use a keyword tool to go deeper. You start with a seed keyword, explore related terms, filter by volume and difficulty, and build out your list from there.
All three methods still work, and they should still be part of the process. But they’re not the process.
Why this is no longer enough
The limitation with each of the methods used to create a keyword strategy is the same: they only show you what's already happening in traditional organic search.
Competitor rankings show you what your competitors have already targeted. If they haven't covered a topic yet, it won't show up. A topic isn’t automatically irrelevant just because a competitor hasn’t covered it.
SERP features like PAA and related searches typically only show relatively short queries. This means you don’t see longer, more specific queries people might type into AI search interfaces.
Keyword databases generally have similar constraints since they're built on organic search data. While metrics like volume and intent are useful for basic filtering, they don’t give you the level of detail you need to create content that truly meets your users’ needs.
These three sources were enough on their own when ranking in organic Google results was all that most marketers cared about. But people increasingly use AI tools to ask more specific questions, and Google search itself now includes AI experiences (like AI Overviews) people can use for more personalized queries that don’t exactly match traditional keywords.
Finally, Google itself is adding new features to account for these changing search behaviors. The company announced at Google I/O 2026 that the search box is expanding to:
- Accommodate longer, more complex queries
- Offer AI-powered suggestions
- Accept input other than text (like images and files)
So the old model misses a growing share of the way people now search for information, products, and brands. This guide will show you how to create a keyword strategy that’s built for this expanded future of search.
What is a keyword strategy?
A keyword strategy is a plan that defines the search queries and AI prompts you want to appear for, how you’ll try to appear for them, and how you want to prioritize them.
A keyword strategy is distinct from keyword research.
Keyword research surfaces data on what people search for, how competitive those terms are, etc. A keyword strategy accounts for that data and your business goals, your resources, and how realistic a given opportunity actually is for you.
For example, keyword research might tell you that “online accounting software” gets searched 10 times more than “accounting software for contractors.” But as part of your keyword strategy, you might decide to focus on the latter because it’s less competitive and more relevant to your audience.
How to create a keyword strategy
Here’s how to build an effective keyword strategy that takes into account your existing visibility, identifies keywords and prompts you should be visible for, and prioritizes them to help you meet your marketing goals:
Step 1. Review your existing visibility
By reviewing your existing visibility in search engines and AI systems, you can identify areas where you’re currently performing well and where there’s room to improve.
Review your visibility using Google Search Console (GSC) and Semrush’s AI Visibility Toolkit.
Your GSC data shows where you already have rankings in Google. Your AI visibility data shows which topics you're already associated with across AI platforms.
When you start identifying new opportunities in the steps below, these baselines will help you separate the topics where you have a foundation to build on from the ones where you're starting from zero. Depending on your authority, resources, and goals, this will help you prioritize which keywords and topics to prioritize.
Check your current Google rankings
To check your current Google rankings, use Google Search Console.
Go to “Performance” > “Search results” and check all the boxes at the top.
Scroll down to the “Queries” table to see the keywords you ranked for over the selected time range, along with the following data:
- Clicks: The number of times users clicked your results
- Impressions: The number of times your results were shown
- CTR: The click-through rate (the percentage of impressions that generated clicks)
- Position: Your average organic ranking position over the selected time range (the lower, the better)
This data gives you a baseline for the individual keywords you already rank for. Click “Export” to save this data in a Google Sheets, Excel, or CSV file.
Check your current AI visibility
Assess your current AI visibility to understand what prompts you’re currently appearing for and where in the AI response you’re showing up.
You can check AI tools manually, but you need to know which prompts to check for. Which is hard to do at scale.
Instead, use Semrush’s Visibility Overview report to see which prompts you appear for. Enter your domain, then select which AI platform you’d like to track using the drop-down at the top.
Scroll to the “Topics & Sources” section to see all the topics your brand appears for. Click into any topic to see the individual prompts you appear for. Then, click “View full response” to see where in the response you show up relative to competitors.
Step 2. Consider your wider marketing goals
Before you start identifying keyword opportunities, you need to know what you're trying to achieve. Because your goals determine which keywords are actually worth pursuing.
The main goals a keyword strategy can serve include:
- Driving organic traffic
- Appearing in AI responses to grow brand awareness
- Generating conversions
- Driving calls and visits for local businesses (see our guide to local keyword research for more on this)
A brand primarily trying to build awareness and show up in AI-generated responses needs broad topic coverage, with a focus on terms relevant to what they do and who they serve.
A brand focused on conversions likely needs to prioritize commercial and transactional keywords that reflect people who are actively evaluating options or are ready to buy.
Decide what success looks like before you start building your keyword list. It changes which opportunities you pursue and how you prioritize them.
Step 3. Identify keyword opportunities
Finding relevant terms you don’t yet rank for and addressing those gaps can help you reach more of your target audience.
Creating high-quality content on topics relevant to your business can also improve your authority in search and AI systems.
There are lots of ways to find keyword opportunities. We detail six methods in our full guide to performing keyword research. But below, we’ll focus on two of the most effective methods to find relevant keywords and prompts.
To track these opportunities, create a spreadsheet with the following columns:
- Keyword/prompt
- Monthly search volume
- Keyword difficulty
- Search/AI landscape
- Search intent
- Competition
- Resources
Add relevant keywords and prompts you find in this step to the “Keyword/prompt” column. We’ll filter and prioritize them later.
Keyword gap analysis
A keyword gap analysis identifies relevant keywords your rivals rank for that you don’t.
Perform your keyword gap analysis with Semrush’s Keyword Gap tool by entering your domain and up to four competitors’ domains, then clicking “Compare.” You’ll see an overview of the keywords that you and your competitors rank for.
Scroll down and select “Missing” to see keywords that all your rivals rank for but you don’t. Or click “Untapped” to see keywords that at least one competitor ranks for but you don’t.
Use the filters at the top to narrow down the list by:
- Position: To only show keywords your competitors appear for in the top 10 search results, for example
- Search volume: To show keywords with a minimum level of search demand
- Keyword difficulty: To filter by your estimated ranking chances
- Intent: To show keywords that align with your goals
You can filter for four broad types of search intent:
- Informational (I): Searchers want to find information
- Navigational (N): Searchers want to find a specific site or page
- Commercial (C): Searchers want to investigate brands, products, or services
- Transactional (T): Searchers want to complete an action (e.g., make a purchase)
When you find a relevant keyword you want to target, add it to your spreadsheet, along with its search volume and keyword difficulty.
Prompt gap analysis
Perform a prompt gap analysis to find prompts your competitors are showing for in AI responses that you’re not with Semrush’s Competitor Research report.
Enter your domain into Competitor Research along with up to four competitors, scroll down to the “Topics & Prompts” table, and select “Prompts.” Click the “Missing” filter to show prompts your rivals appear for that you don’t.
Add each prompt you want to track to your spreadsheet.
Since it’s an AI prompt rather than a keyword with typical search volume and keyword difficulty metrics, just add “N/A” to the relevant columns. You know there’s demand for these terms as they show up in the tool, and we’ll analyze how likely it is that you’ll appear for them later on.
Step 4. Consider the search landscape
Analyzing the search landscape for each keyword helps you understand how likely users are to see and click your page to help you determine whether the keyword is likely to drive results for your brand.
Start by searching each keyword in Google and noting what appears before the first organic result. Paid ads, shopping carousels, AI Overviews, map packs, and featured snippets all tend to push organic results farther down the page and reduce the chances a user will click on them.
Analyzing the Google SERP like this helps you understand the likelihood of your content driving traffic. But it can also help you determine what kind of page you might need to create to appear for that keyword.
For example, paid ads and shopping results suggest you’ll need a product page. While a YouTube carousel at the top of the SERP could indicate a video might perform better than a blog post.
For AI prompts, type them into tools like Google AI Mode or ChatGPT to perform a similar analysis but for the AI search landscape instead. In particular, note things like:
- Whether the AI tool recommends specific brands or products
- What kinds of pages are being cited as sources (e.g., blog posts or videos)
- Whether the AI tool displays product or shopping cards within the response
Note any findings in the “Search/AI landscape” column in your spreadsheet.
Step 5. Evaluate search intent
Evaluating search intent (what the searcher is trying to achieve) helps you create content tailored specifically to their needs, which means it’s more likely to appear in search results and AI responses.
For example, you’d probably struggle to rank high for “nintendo switch 2” with a blog post. The search results are dominated by product listing ads and product pages. This makes sense since a user typing in this query is likely looking for the product and not an informational post.
Intent can also tell you where the user is in the marketing funnel, and how likely they are to convert. This can help you map keywords to your goals.
For example, someone who searches “nintendo switch 2 games” is more likely to make an immediate purchase than someone who searches “upcoming nintendo switch 2 games.”
Understanding search intent goes beyond simple categories. It requires looking deeper to find the more nuanced intents behind the keywords you’re targeting. This can help you understand what kind of content you should create for each keyword.
When you search a keyword in Google or an AI tool, look at what's ranking and/or being cited and note three things:
- The format of the content (e.g., guides, listicles, product pages, or videos)
- The angle most top results or cited sources take (like beginner-focused or expert-level)
- The depth they go into
You shouldn’t match what competitors do exactly. But trends here can indicate the kind of content search engines and AI tools reward — and what your target audience is looking for.
Add your findings to the “Search intent” column in your spreadsheet.
Step 6. Assess your ability to compete
Analyzing how difficult a keyword would be to appear for helps you prioritize terms you have the best chance of getting results from.
You can get a quick gauge of a keyword’s difficulty using a tool like Semrush’s Keyword Overview. You can even add your domain to get a Personal Keyword Difficulty score:
But to get a true idea of how difficult a term will be to appear for, you need to analyze the competition. Do that by scrolling down within Keyword Overview to the “SERP Analysis” section to see the Authority Scores (i.e., overall quality and search strength scores), backlinks, and estimated search traffic of each domain ranking for that term.
If most of the top results have a lot more authority and backlinks than you do, that keyword may be difficult to rank for without first increasing your own authority.
Aside from understanding authority, assess whether you need extensive resources to create something comparable to competing pages. Note anything that stands out in the spreadsheet’s “Competition” column.
For example, do your rivals perform extensive product testing? Do the articles contain original data you’d struggle to replicate? Do any videos that appear have higher production value than you can manage?
Step 7. Take stock of your resources
Take stock of your resources to understand how you can effectively implement your keyword strategy.
This includes your (and your team’s):
- Time
- Money
- Skills
- Knowledge
- Tools
For example, you’ll struggle to rank for “how to play pickleball video” if you don’t have the skills and tools required to make a video. Or the budgetto hire a professional to record one for you.
And if the top-ranking (or top-cited) terms for a given query are pieces of original research, you’ll struggle to compete if you don’t have the staff or data to create your own study.
It’s also helpful to take stock of your existing assets, such as:
- Blog posts
- Ebooks
- Videos
- Testimonials and case studies
- Proprietary data
- Images
A keyword might not need a completely new piece of content if you already have a video or ebook you could repurpose into a blog post or another format.
Add any requirements or existing assets to the “Resources” column in the spreadsheet.
Step 8. Organize your keywords
Organize your list of keywords and prompts by sorting them into topic clusters, which are groups of thematically related and interlinked pages that establish deeper authority in particular areas related to your business.
Each topic cluster includes two types of pages:
- Pillar page: Think of a pillar page as the main hub or foundation for a broad topic
- Subpages: These are more specific pages that dive into subtopics of your pillar page’s main topic
Thinking in terms of topics rather than individual search terms also mimics how AI tools respond to user queries. For example, someone typing “how to do email marketing” into Google might get an AI Overview response like this:
Notice how the AI Overview covers the topic broadly, including sections on:
- How to choose a platform
- How to grow your email list
- Segmenting your audience
- Writing good emails
- Automation and tracking
Structuring your keyword strategy around topics rather than individual keywords can help you appear in more of these responses. Because it leads to naturally creating content that covers and links to related subtopics.
To cluster your content, download your spreadsheet as a CSV and upload it to your favorite AI tool (I find Claude works best for this) and enter this prompt:
“Attached is a CSV of keywords and prompts related to [main topic], along with metrics and notes for each one. Cluster them into topic groups based on shared themes and search intent.
For each cluster, identify one pillar topic (the broadest term that represents the theme) and the keyword(s) and/or prompt(s) it should target as well as the subpages below it for the remaining keywords.
Output the results as a PDF that organizes all the keywords and prompts into pillar pages and associated cluster pages. Include the keywords for each cluster, one per line, as plain text (not bullet points). Add the relevant notes alongside these.”
You should now have a PDF with a list of pages and the context you added for each one, which you can use or send to your content team to start creating the relevant content yourself.
Here’s an example output:
Step 9. Track your results and keep adapting
Monitor results from implementing your keyword strategy by tracking your visibility for target keywords and prompts.
With Semrush’s Position Tracking tool, you’ll see where you rank for your target keywords and get alerts about important changes. Just click on the bell icon after setting up your project and set the desired trigger(s).
Then, go to the “Overview” tab and scroll down to the “Rankings Overview” table to see rankings for your target keywords.
To track your AI Overview visibility, use the “SERP Features” filter, select “[yourdomain].com ranks,” and choose “AI Overview” as the SERP feature. This will show your visibility for keywords you appear in the AI Overview for.
You can also use Position Tracking to track prompts you want to appear for in platforms like ChatGPT, Gemini, and Google AI Mode. To do this, set up your project by choosing one of the AI platforms under “Search engine,” choose a location, and click “Continue To Prompts.”
Once you’ve added your prompts, go to the “Overview” tab to see whether you’re mentioned for each prompt, along with changes in your position and visibility.
Use this data to identify which keywords and prompts you’re gaining or losing traction for. Look for issues with your content that could be holding you back, or assess the competitive landscape to see if the keyword or prompt may no longer be achievable.
Build your own keyword strategy
Creating an effective keyword strategy is key to producing content that your audience is looking for and that helps you meet your goals.
By prioritizing and targeting the right keywords and prompts, you can improve your chances of ranking well in search engines and appearing in AI responses.
Sign up for a Semrush One free trial today to access tools to help build your own keyword strategy.