AI & Automation
Personas
SaaS & Startup
Time to ROI
Short-term (< 3 months)
Last month, I took on a B2B startup website project where the first critical step was obvious: build a comprehensive keyword list that would actually drive qualified traffic. Like most SEO professionals, I started where everyone begins—firing up SEMrush, diving into Ahrefs, and cross-referencing with Google autocomplete.
After hours of clicking through expensive subscription interfaces and drowning in overwhelming data exports, I had a decent list. But something felt off. The process was expensive (multiple tool subscriptions adding up), time-consuming (endless manual filtering), and frankly overkill for what I actually needed.
That's when I discovered something that changed my entire approach to keyword research. What started as frustration with traditional tools led me to experiment with AI-powered research methods that not only saved money but delivered better, more contextual results.
Here's what you'll learn from my real-world experiment:
Why traditional SEO tools often overcomplicate keyword research
The specific AI workflow I developed that replaced $200+/month in tool subscriptions
How AI understands search intent better than keyword volume metrics
The exact prompts and processes I use for different types of keyword research
Why this approach works better for startups and growing businesses
This isn't about replacing all SEO tools—it's about building a smarter, more cost-effective research process that actually fits how most businesses operate. Here's exactly how I did it.
Industry Reality
What every marketer has been told about keyword research
The SEO industry has been preaching the same gospel for years: you need expensive tools to do proper keyword research. Every course, every guru, every agency tells you the same story.
The traditional keyword research playbook goes like this:
Subscribe to multiple tools (SEMrush, Ahrefs, Ubersuggest, etc.)
Export thousands of keyword suggestions
Filter by search volume and keyword difficulty
Cross-reference competition data
Create massive spreadsheets with hundreds of keywords
This approach exists because, historically, we needed these tools to access Google's search data. The logic was simple: more data equals better decisions. But here's what nobody talks about—most of that data is either inaccurate or irrelevant for small to medium businesses.
The problems with traditional keyword research:
Search volume data is notoriously unreliable (tools often show 0 searches for keywords that actually drive 100+ visits monthly)
You're paying for features you'll never use (enterprise-level competitive analysis for a startup?)
The process is overwhelming—who has time to analyze 10,000 keyword suggestions?
Tools focus on volume over intent, missing the keywords that actually convert
Most businesses end up with analysis paralysis, spending more time researching keywords than actually creating content. The industry has convinced us that complexity equals quality, but I learned that's not always true.
Consider me as your business complice.
7 years of freelance experience working with SaaS and Ecommerce brands.
When I started working on this B2B startup project, I initially followed the traditional playbook. I fired up SEMrush, exported keyword lists, and started the familiar dance of filtering and analyzing. But after spending hours drowning in data that felt disconnected from the actual business, I hit a wall.
The startup was in a specialized niche—they provided workflow automation software for small accounting firms. The keyword tools kept suggesting generic terms like "accounting software" and "workflow management," but these felt too broad and competitive for a startup with no domain authority.
My first approach failed because:
The suggested keywords were either too competitive or too generic
Volume data showed zeros for terms I knew their customers were searching for
I was spending more time in tools than understanding the actual business
That's when I remembered something: I had a dormant Perplexity Pro account. On a whim, I decided to test their research capabilities for keyword work, mainly because I was frustrated with the traditional approach.
I started with a simple query: "What specific problems do small accounting firms face with workflow automation?" The response was immediately different from what SEMrush would give me. Instead of a list of keywords with volume numbers, I got contextual insights about real problems, pain points, and the language people actually use when discussing these issues.
That's when it clicked. AI tools don't just find keywords—they understand context, search intent, and the relationship between problems and solutions in ways that traditional tools miss.
The difference was immediate and shocking. What would have taken days of keyword research turned into hours of strategic understanding.
Here's my playbook
What I ended up doing and the results.
After that initial breakthrough with Perplexity, I developed a systematic approach that completely changed how I do keyword research. This isn't about completely abandoning traditional tools—it's about using AI to make the entire process smarter and more strategic.
My 4-Step AI-Powered Keyword Research Workflow:
Step 1: Problem-Intent Mapping
Instead of starting with seed keywords, I start with customer problems. I use AI to understand the language customers actually use when describing their pain points.
My go-to prompt: "What are the specific problems [target audience] face with [broad topic]? List the exact phrases and terminology they use when searching for solutions."
For the accounting firm client, this revealed search terms like "automate client onboarding for small CPA firms" and "reduce manual data entry in accounting workflows" - terms that never appeared in traditional keyword tools but were exactly what their customers needed.
Step 2: Competitive Context Research
I use AI to analyze what competitors are actually ranking for and why. Instead of just seeing a list of keywords, I get insights into content gaps and opportunities.
Prompt: "Analyze the content strategy of [competitor websites]. What topics are they covering well, and what gaps exist that a [business type] could fill?"
Step 3: Search Intent Clusters
Traditional tools group keywords by volume. I group them by intent and customer journey stage using AI analysis.
Prompt: "Group these search terms by user intent and buying stage: [list of terms]. Explain why someone would search for each group and what content would best serve that intent."
Step 4: Content-Keyword Alignment
Finally, I use AI to map keyword opportunities to specific content types and business goals.
The result? Instead of a spreadsheet with 500 random keywords, I had a strategic map of exactly what content to create and why. Every keyword suggestion came with context about who would search for it, when in their journey, and what type of content would convert them.
This approach helped me build the client's entire content strategy in a fraction of the time it would have taken with traditional tools.
Strategic Foundation
Start with customer problems and pain points, not search volumes. AI excels at understanding the language customers actually use.
Intent Over Volume
Focus on search intent and customer journey mapping rather than chasing high-volume competitive terms that won't convert.
Contextual Discovery
Use AI to find keyword opportunities that traditional tools miss because they understand semantic relationships and context.
Validation Method
Combine AI research with small-scale content testing to validate which keywords actually drive qualified traffic.
The impact of switching to this AI-powered approach was immediate and measurable. Within the first month of implementing this workflow, I had completely changed how I approach keyword research for all my clients.
Immediate Results:
Reduced keyword research time from 2-3 days to 4-6 hours per project
Cut monthly tool subscriptions from $200+ to $20 (just Perplexity Pro)
Identified 40+ high-intent keywords that traditional tools had missed completely
Built a content strategy that actually aligned with customer language and needs
For the accounting firm client specifically, this approach helped us identify niche terms that their actual customers were using. Instead of competing for "accounting software" with enterprise companies, we targeted specific workflow problems that small firms face.
The most surprising outcome? The keywords AI suggested often had "zero" search volume according to traditional tools, but when we created content around them, they drove consistent, qualified traffic. This confirmed what I suspected—volume data from traditional tools is often wrong, especially for niche B2B terms.
Six months later, I've used this approach for over a dozen projects across different industries, and the pattern holds: AI-powered keyword research consistently finds opportunities that traditional tools miss.
What I've learned and the mistakes I've made.
Sharing so you don't make them.
After implementing this approach across multiple client projects, here are the key insights that changed how I think about keyword research:
1. Context beats volume every time. A keyword with "zero" searches that perfectly matches customer language will outperform a high-volume generic term.
2. AI understands semantic relationships.** Traditional tools see "workflow automation" and "process optimization" as different keywords. AI understands they're often the same search intent.
3. Customer language evolves faster than keyword tools.** AI picks up on new terminology and phrases that haven't made it into traditional databases yet.
4. Intent matters more than competition.** I'd rather rank #3 for a high-intent keyword than #1 for a keyword that doesn't convert.
5. The research process should inform content strategy, not just SEO.** AI-powered research naturally leads to better content ideas because it focuses on problems and solutions.
6. Most businesses need fewer, better keywords.** Instead of targeting 500 keywords poorly, focus on 50 that perfectly match your customer's journey.
7. Traditional tools are still useful for validation.** Once AI identifies opportunities, traditional tools can help validate commercial potential and competition levels.
The biggest lesson? Stop treating keyword research as a data collection exercise and start treating it as customer research. That's where AI really shines.
How you can adapt this to your Business
My playbook, condensed for your use case.
For your SaaS / Startup
For SaaS startups implementing this approach:
Focus on problem-solution keywords rather than feature-based terms
Use AI to understand the language your ICP actually uses in their day-to-day work
Map keywords to different stages of the customer journey from problem-aware to solution-aware
Prioritize long-tail keywords that indicate buying intent over broad competitive terms
For your Ecommerce store
For ecommerce stores using this method:
Research product-specific pain points and use cases rather than just product names
Use AI to discover seasonal and trending search behaviors in your niche
Focus on "buying intent" keywords that indicate someone is ready to purchase
Combine AI research with customer reviews to find the exact language buyers use