What I Do Before Writing Any Prompt

Figure Out My Goal

I ask myself: what do I want from this content?

• More website visitors?

• People buying something?

• Answering common questions?

Then I think about what people are trying to do when they search:

• Learning something (informational)

• Buying something (transactional)

• Finding a specific site (navigational)

Know My Audience

Who are these people? What problems do they have? How do they talk?

If I'm writing for small business owners, they might search "how to get more customers" not "customer acquisition strategies."

Pick My Starting Topics

I need seed keywords. These are the main topics I already know about or want to cover.

I get these from:

• Content I already have

• What competitors write about

• Just brainstorming obvious topics

Set the Rules

Before I start, I decide:

• How much search volume do I want?

• How hard is too hard to rank for?

• Am I targeting local or global searches?

• What topics should I avoid?

How I Write My First AI Prompt

A good prompt needs these things:

1. My main topic plus any subtopics

2. What kind of keywords I want (long phrases, questions, etc.)

3. Volume and competition limits

4. What people are trying to do when they search

5. Location or language details

6. How I want the results formatted

Here's an example I might use:

"Generate 20 long-tail keywords with decent search volume and low competition about email marketing for small businesses. Focus on informational and how-to searches. Skip any branded terms. Group them by what people are trying to do."

Why this works:

• I'm specific about the topic

• I tell it what kind of keywords I want

• I mention competition level

• I explain the search intent

• I ask for organization

Being vague gets you generic results. Being specific gets you useful keywords.

How I Make My Prompts Better

I run my first prompt and look at what I get. Then I ask follow-up questions like:

• "Give me more specific versions of these"

• "What keywords are my competitors using that I missed?"

• "Which of these are easiest to rank for?"

Sometimes I try the same prompt with different wording to see what changes. I might combine results from multiple attempts.

The key is not settling for the first results. Keep pushing for better keywords.

Checking If the Keywords Are Actually Good

AI gives me ideas but I need real data. I use tools like:

• Ahrefs

• SEMrush

• Google Keyword Planner

I check:

• Actual search volume

• How hard it'll be to rank

• What's already ranking for these keywords

I also ask AI to help analyze: "From this list, which 10 keywords should I target first based on difficulty and volume?"

Google Trends shows me if a keyword is getting more popular or dying out.

What I Do With the Keywords

I group similar keywords together. Like all the "how to" keywords go in one pile, all the "best" keywords in another.

Then I plan content around these groups. Each group might become one blog post or page.

I pick which keywords to work on first. Usually that's:

• Lower competition

• Decent search volume

• Highly relevant to what I do

When I write, I use these keywords in my titles, headings, and content naturally.

Mistakes I Made Early On

Being too vague: I'd ask for "SEO keywords" and get generic junk. Now I'm super specific about my topic and audience.

Only trusting AI: The AI doesn't know real search volumes or competition. I always verify with actual SEO tools.

Ignoring search intent: I'd target keywords without thinking about what people actually wanted. Big mistake.

What I learned: always double-check everything. Mix AI creativity with real data and human judgment.

Real Example

Let's say I'm targeting "productivity apps for remote workers."

My prompt: "Generate 15 long-tail, low-competition keywords about productivity apps specifically for remote workers. Focus on comparison and recommendation searches. Exclude specific app names. Show search intent for each."

AI gives me things like:

• "best productivity apps for working from home"

• "remote work apps vs in-office tools"

• "productivity software for distributed teams"

I check these in Ahrefs. Some have good volume and low difficulty. Others are too competitive.

I pick the best 5 and plan content around them. Maybe one becomes "The 7 Best Productivity Apps for Remote Workers" and another becomes "Remote Work Apps: What You Actually Need."

Tools I Use

AI tools:

• ChatGPT

• Claude

• Gemini

• Sometimes DeepSeek

SEO tools:

• Ahrefs (my main one)

• SEMrush

• Google Keyword Planner for basic volume data

Free stuff:

• Google Trends

• Reddit for seeing how people actually talk

• Industry forums

I keep templates of prompts that work well so I don't start from scratch every time.

Wrapping Up

My process is simple:

1. Set up my goals and constraints

2. Write a specific prompt

3. Keep improving the results

4. Verify everything with real data

5. Actually use the keywords in content

This approach is way faster than old-school keyword research. And I find keywords I never would have thought of on my own.

Your prompts will get better as you practice. Don't expect perfect results right away. Also, search behavior changes. Come back and do fresh keyword research every few months to stay current.