First Principles of Keyword Research For Content Marketing and SEO

Keyword research isn’t a list-building task. It’s positioning. It’s business insight with a search bar. I don’t start with tools. I begin with truths — customer pain, product utility, and market potential.

1. I Start With Business, Not Keywords

Before I touch a tool, I ask:

• What exactly does the product do?

• Who is it for, and why do they care now?

• What moment of pain or desire does it serve?

Example: If I’m working on a meditation app, I don’t begin with “guided meditation.”I think in moments:

“When I feel anxious and can’t sleep, I want something to calm my mind, so I can function tomorrow.”

That’s the seed. Everything else grows from there.

2. I Use “Jobs To Be Done” as Step Zero

This is always where I start, especially in new markets.

JTBD Template:

“When I [situation], I want to [do/feel/solve], so I can [achieve/avoid outcome].”

Examples:

• Mental Health: “Why does no one understand me?”→ lonely, talk to someone, breakup recovery

• Career Switch: “I hate my job.”→ Top courses, high-paying skills, what to do after an MBA

The goal is to reverse-engineer the why behind the search. That’s where real keyword intent lives.

3. I Expand Seed Keywords Using 3 Lenses

A. Product Features

I list out every use case, feature, and micro-problem the product solves.

Example: A video tool =→ crop video, add music to video, merge clips online

B. Real-World Language

Users don’t always speak “product.” They speak of “pain.”

I go to Reddit, forums, support tickets, and app reviews to find synonyms and emotional language.

Examples:

• “Alone” → abandoned, ignored, left out

• “Productivity” → focus, discipline, procrastination hacks

C. Micro-Moments (Google’s Term)

These are situational triggers that spark high-intent queries.

• Lost charger → buy a charger near me now

• Breakup → How to stop crying after a breakup

If I get these right, I unlock the user’s actual intent, not just the query.

4. I Use Tools, But Only After the Thinking

Here’s what I use — and when:

• Ahrefs → Matching terms, content gaps, and reverse-engineering competitors

• Ubersuggest → Cheaper starter tool

• Keyword Insights → Clustering by intent (article vs tool vs product page)

• Google Suggestions → PAA, Autocomplete, Related Searches

• Forums & Reddit → Emotionally charged, long-tail gems

• Support Logs & Reviews → Real language that no tool will catch

Rule: Tools support my thinking. They don’t replace it.

5. I Build a Keyword Universe — Not Just a List

I don’t stop at keywords. I map the entire journey.

Example: Meditation App

Stage	Keywords
Aware	Why am I always stressed? How to stop overthinking
Consideration	best meditation app for beginners, sleep meditation guide
Decision	Calm vs Headspace, BlockerX reviews

I cluster by parent topics, journey stages, and user state.

This isn’t SEO. This is user empathy at scale.

6. I Prioritize Ruthlessly with the P.I.E. Framework

I score keywords using:

P.I.E. = Potential × Intent × Effort

A. Potential

• Volume (yes — but avoid vanity)

• Business value (can this lead to revenue?)

• Authority potential (can I “own” this space?)

B. Intent

• Informational → Blog/Video

• Transactional → Landing page/Tool

• Navigational → Brand, reviews, alternatives

I validate with SERP analysis or Keyword Insights clustering.

C. Effort

• Keyword Difficulty (Ahrefs)

• Domain Rating (DR) — Can we realistically rank?

• Resources — Content, Dev, Design… who’s available?

7. I Clean Out the Junk

I actively reject keywords that don’t align with product or revenue.

I Remove:

• Irrelevant terms (e.g., “Madhuri Dixit movies” for a mental health site)

• Informational-only queries (if I’m selling a tool, not blogging)

• Intent mismatches (if the user wants a product, but my page is an article)

Don’t fall for volume traps. If your product can’t serve the query, toss it.

8. I Use Both Top-Down and Bottom-Up Approaches

I switch between two modes:

Bottom-Up (Scrappy Mode)

• Start with what we already rank for

• Build depth → internal links, upgrades, FAQs, interlinking

Top-Down

• Map the entire market (TAM, competitors, gaps)

• Reverse engineer winners

• Plan content and links from the outside-in

“We don’t do SEO for rankings. We do it to capture market whitespace.”

9. I Deliver These Two Docs

A. Keyword Matrix Sheet

• Keyword

• Volume

• KD

• Intent

• Suggested format

• Priority score

B. SEO Opportunity Sheet

• TAM (Traffic Opportunity)

• DR gap

• Number of articles/backlinks required

• Aggressive vs Conservative path

I use these to align with founders, PMs, and marketers. No fluff. Just math.

10. 🤖 Where I Use AI — And Where I Don’t

AI Helps Me:

• Generate long-tail variations

• Cluster keywords by intent

• Summarize support transcripts

AI Can’t Do:

• Finalize seed list (only the user context can)

• Prioritize high-value terms

• Understand the business context or buying triggers

Mantras I Live By

“Keyword research is business research — just with a search box.”“If the user doesn’t care, the keyword doesn’t matter.”“Your ship. My compass. Our glory.”

Keyword Research Examples and Field Notes

I. User Context & Customer Research

A. Catching Language Mismatches

• “Integration” vs “Sync”: In my first job at a data integration company, I kept saying “integrate apps.” But in just 4–5 user calls, I realized everyone said “sync data.” Keyword tools didn’t show it — customers did. I rewrote the copy and content plans around “sync.”

• Pricing Page Blindspots: At a SaaS startup, we kept optimizing around “pricing” and “plans.” But user feedback showed queries like “how much does X cost?” and “cheapest way to do Y.” That triggered a full content pivot.

B. Getting Closer via Support, Surveys, and Interviews

• Customer Support > Keyword Tools: I’ve often started keyword research by reading support tickets or sitting in on onboarding calls. I pull transcripts into a word cloud, look for repeated phrasing, and build keyword pools from there.

• Prana Care (Mental Health): Through surveys, we found that users weren’t choosing us because of our qualifications — they just wanted a simple onboarding process. So “easy mental health access” and “talk to someone fast” became seed phrases.

• Lido Learning (EdTech): We conducted conversion interviews to understand why users chose us over our competitors. Questions like “what 3 websites did you compare before buying?” helped surface terms our competitors ranked for — but we didn’t.

C. Listening Beyond the Website

• App Review Mining: For Misho (a reseller app), I analyzed one-star reviews to identify key phrases such as “delivery failed,” “payment bug,” and “order not accepted.” These weren’t just product issues — they were hidden content opportunities.

II. Business Context & Alignment

A. Product-First Keyword Thinking

• Practo: SEO was the product. We deployed hundreds of on-ground agents to collect verified clinic and doctor data. That hyper-local content made us unbeatable for terms like “dentist in Bangalore.” The SEO edge came from data no one else had.

• Prana Care: The CEO gave us four pillars — productivity, habits, discipline, and wellness. My job was to break each into micro clusters like “habit tracker for students,” “how to build discipline.” That’s how we built topical authority in narrow lanes, not broad oceans.

B. TAM and the Case for SEO Investment

• Edureka: I ran a top-down TAM analysis and pitched a plan to leadership. The ask: ₹1.2 Cr. I showed what 3.8L visits/mo would take, in terms of content, links, and time. I gave two paths: conservative and aggressive. They chose aggressive. And we hit targets.

• SaaS Keyword Planning: I always involve founders and product leads. in this process I ask: “What do your best users Google before they find you?” That insight is more valuable than anything Ahrefs produces.

III. Turning JTBD Into Seed Keywords

A. Real Jobs → Real Queries

• “I feel lonely.”→ how to deal with loneliness, nobody understands me→ Seed terms: lonely, abandoned, alone, friendless

• “I want to relax”→ tight chest, stress, how to calm anxiety, Ayurvedic massage

  That’s not a content calendar — that’s emotional relevance at work.

B. Expanding Seeds via Synonyms

• Used Cars: We missed major traffic from people searching for second-hand and refurbished, not just used.

• Design Ops (Early Niche SaaS): Search volume was low. However, I identified adjacent problems (such as team onboarding and design feedback loops) and discovered a broader, untapped universe.

IV. Mapping the Keyword Universe

A. Using TAM → SAM → SOM to Model Keyword Scope

• Prana Care & Edureka: I used this mental model:

  • TAM = “I want to feel better”

  • SAM = “talk to therapist”

  • SOM = “online therapy for students”

Then I mapped seed → cluster → landing page and back-calculated the effort required to rank.

• Edureka’s Full Journey: From what is HTML to DevOps resume 5 years experience — I plotted it all. That’s how we uncovered both early intent traffic and job-ready purchase queries.

V. Prioritizing or Rejecting Keywords

A. Not All Keywords Deserve Content

I use a formula: Keyword Difficulty × Search Volume × Business Fit × DR × Conversion Potential = Final Score.

I’ve rejected high-volume terms like “Madhuri Dixit movies” for a mental health site. Why? Because they don’t bring buyers. I’ve killed hundreds of “content for traffic’s sake” projects this way.

Tools I Use (and How)

• Ahrefs: To reverse-engineer what top competitors rank for, and break down their best subfolders

• Keyword Insights: For clustering by intent (blog vs product vs comparison)

• Search Console + Data Studio: For measuring non-branded lift and week-by-week content ROI. Tools come later. First comes thinking.