- Tim at Penfriend
- Posts
- POV: Your keyword research is 12 words holding hands
POV: Your keyword research is 12 words holding hands
How one question changed how I do keyword research. Forever.

Day 225/100
Hey—It's Tim.
I’ve been thinking about a conversation I had a few weeks ago with a CMO.
It feels like keyword research is just rearranging the same 12 words over and over
Ooof. I felt that.
For the longest time, I did the same thing.
Started with what I already knew: old blog tags, terms we ranked for, phrases people used in Slack. Super comfortable. Also super narrow. It's like fishing off the same dock every single day and then being shocked when you never catch anything new.
The thing that changed everything for me wasn't a tool. It was a question:
“How does someone's intent change as they learn more?”
Once I started mapping that learning curve, my clusters stopped looking like synonyms holding hands in a circle. They started looking like actual routes people take.
Let me show you exactly what I do now.
Let's use a real example
I'm going to walk through one example the whole way, so you can see how this actually works.
Let's say you're working on content for a project management tool.
You've got a prospect—let's call her Maya.
She's a RevOps lead at a 120-person SaaS company. Right now, her team is drowning in weekend spreadsheet merges. Every Friday at 5pm, someone manually pulls data from Salesforce, HubSpot, and their support tool, then spends two hours combining everything in Google Sheets. Every. Single. Week.
That's the job to be done: "Replace weekend spreadsheet merges with a button."
That's your hub. That one sentence. Not "improve data workflows" or "enhance operational efficiency" or any of that fluffy nonsense.
The actual human problem: weekend spreadsheet merges that need to die.
Here's how I find the spokes
Maya doesn't wake up one morning and immediately buy your tool. She goes through a whole journey. Let's map what she actually does:
First, she tries the duct tape solution:
She sets up a Zapier chain to auto-merge the data
It works! For three weeks.
Then it starts failing at 2am on Saturdays
She's getting Slack messages all weekend: "Where's the report?"
Then she explores real solutions:
She starts googling "Zapier alternatives for data syncing"
She asks in a RevOps Slack: "Anyone solved HubSpot + Salesforce reporting without dying?"
She downloads three tools and realizes IT will block all of them
Then she hits the internal wall:
She finally finds a tool that might work (yours!)
But now she needs to convince the CFO
Who wants "proof of three-month payback"
And the IT director who wants to see the security docs
Each of those moments? That's a spoke.
See how we went from one job statement to a whole map of content you need to create?
Now let's build the actual cluster
Here's what I do next. I open up ten search results for each phase Maya's going through.
Let's start with that early phase: "Zapier alternatives for data syncing"
I'm not reading the articles. I'm collecting patterns.
I pull out three types of things:
Entities (the proper nouns):
Zapier, Make, Workato, Tray.io
HubSpot, Salesforce, Stripe
CSV, API, webhook
RevOps, Finance, IT
Formats (how the content is shaped):
"Top 10" lists
Comparison tables
Templates
Calculators
Checklists
Video teardowns
Stages (where they are in decision-making):
"vs" (Zapier vs Make)
"alternatives to"
"for [role]" (for RevOps teams)
"pricing"
"how long does [thing] take"
"migration guide"
Now here's the fun part. You mix and match these.
Entity × Format × Stage
So "Zapier alternatives for data syncing" explodes into:
"Zapier vs. Make vs. Workato for RevOps teams: 2025 comparison table"
"HubSpot + Salesforce sync calculator: ROI in hours saved per month"
"Data sync migration checklist (with rollback plan)"
"How long does Zapier to Workato migration take? Real timelines from 5 RevOps teams"
Start with 2 of the variables against each other, and as you cross all those, then make the 3’s. Better yet, just add the 3’s as sections into the already created blog.
You're not guessing anymore. You're combining what the market's already teaching buyers to expect.
Make every spoke lead somewhere real
Here's where most content dies: it answers a question but gives you nothing to hold onto.
Maya googles "How to sync HubSpot and Salesforce without Zapier"
Most content gives her a 2,000-word essay about integration methods. Cool. Useless.
What she actually needs: a thing she can use right now.
So instead, you create:
"HubSpot + Salesforce sync checklist (with field mapping template you can copy)"
The template is:
{question} → {deliverable} + {constraint}
More examples from Maya's journey:
"Why does my Zapier chain keep failing?" → "Zapier failure debug flowchart (5-minute diagnosis)"
"How to convince CFO to replace Zapier" → "Three-month payback calculator with ready-to-send CFO memo"
"What's the fastest way to migrate off Zapier" → "Weekend migration runbook (with rollback script)"
Write the deliverable first.
If you can't name an actual file you'd attach to an email - a template, a checklist, a calculator - the spoke is probably too mushy. Keep tightening until you can picture dragging a .xlsx into Slack.
Put this into practice (today)
Okay, let's make this concrete. Here's a 90-minute sprint you can run this afternoon.
Minutes 0-20: Find your hubs
Write down three recent wins or losses.
For each one, extract the job in plain language:
"Replace weekend spreadsheet merges with a button"
"Get CFO sign-off without a three-month pilot"
"Cut report building from 4 hours to 20 minutes"
Those are your hubs.
Minutes 20-45: Map the journey
Pick one hub. Write down the six things someone does before and after that job gets solved.
For Maya:
Sets up Zapier (it breaks)
Asks in Slack for alternatives
Downloads three tools
Hits IT security wall
Needs CFO business case
Needs migration plan with safety net
Those six things are your first six spokes.
Minutes 45-70: Gather the patterns
Open ten search results for "Zapier alternatives."
Pull out:
Entities: Zapier, Make, HubSpot, Salesforce, RevOps
Formats: comparison table, calculator, checklist, template
Stages: "vs," "for RevOps," "migration," "pricing"
Minutes 70-90: Build six pieces
Combine entity × format × stage for six content ideas:
"Zapier vs. Make vs. Workato for RevOps: 2025 comparison"
"HubSpot + Salesforce sync calculator (ROI in hours saved)"
"Data sync migration checklist (with rollback plan)"
"Weekend reporting template for 50-200 person teams"
"CFO business case: Data automation payback calculator"
"IT security questions: Pre-built answers for sync tools"
Each one needs a deliverable. A file. Something they can download or copy.
End of day: Ship one
Pick the easiest spoke. Write it. Publish it.
Don't wait for perfect.
Edit tomorrow.
The whole map for Maya
Let me zoom out and show you what the full cluster looks like:
Hub: "Replace weekend spreadsheet merges with a button"
Early exploration spokes:
Zapier vs. Make vs. Workato comparison (with RevOps-specific scoring)
"Why does my Zapier chain keep failing?" debug flowchart
Community thread: "Best Zapier alternatives for HubSpot + Salesforce"
Decision spokes:
CFO business case template with three-month payback proof
Migration runbook (with rollback script if things go wrong)
"How long does migration actually take?" (Real timelines from five teams)
Evaluation spokes:
Data sync ROI calculator (hours saved per month)
Security checklist for IT approval
Field mapping template (the actual spreadsheet they can use)
Format variety across the same intent:
The comparison article
The calculator
The checklist
The template
The case study
The video teardown
You're not just covering synonyms. You're covering the entire route someone takes from "I have a problem" to "I bought a solution."
I have 8 more go to ways on how I build out clusters. This is just a current favourite.
Do you want the rest?
Give it a shot and let me know how it goes? I'm curious if the Maya example makes sense or if I should walk through a different type of buyer journey.
See you tomorrow.

✌️ Tim "Tabs Open = % Locked In" Hanson
CMO @Penfriend.ai
Same brain, different platforms: X, Threads, LinkedIn.
P.S. One more thing
Keep an anti-keyword list.
For Maya's cluster, you'd exclude:
"influencer"
"B2C"
"small business"
"solopreneur"
"side hustle"
This keeps your content tight.
You're not trying to rank for everything.
You're trying to own one specific route.

Penfriend.ai
Made by content marketers. Used by better ones.
What to do next
Share This Update: Know someone who’d benefit? Forward this newsletter to your content team.
Get your First 3 Articles FREE We just dropped the biggest update we’ve ever done to Penfriend a few weeks ago.
Let Us Do It For You: We have a DFY service where we build out your next 150 articles. Let us handle your 2025 content strategy for you.