Cristian Tala

Posted on Mar 12

Weekly SEO Research on Autopilot (Saving 12 Hours/Month)

#seo #automation

I do SEO research every Monday at 1 AM. Not manually — a cron job does it for me.

Output:

Markdown report with keyword gaps
7 tasks created in NocoDB with priority
E-E-A-T analysis (which keywords leverage my exit/experience)

Time invested: 5 minutes reading the report (vs 3 hours/week manually).

Savings: 12 hours/month.

It's not magic. It's a system that automates the tedious stuff (fetching data, comparing lists, detecting gaps) so I can focus on the strategic work (what to write, how to position it).

Here's how it works (with code included).

The Problem: Manual SEO Is Tedious

Traditional SEO research:

Open Ahrefs/SEMrush/Serpstat
See keywords you rank for
See keywords competitors rank for
Compare lists manually (Excel hell)
Identify gaps ("they rank for X, I don't")
Prioritize by volume + difficulty
Create tasks in task manager

Time: 2-3 hours/week.

Frequency: Weekly (because keywords change, competitors publish).

Result: SEO gets deprioritized because it's boring.

The Solution: Automated SEO with Serpstat API

My setup:

API: Serpstat (keyword research + competitor analysis)
Cron: Monday 1 AM (isolated session, no clutter)
Model: Opus 4.6 (strategic analysis requires deep thinking)
Output: Markdown report + NocoDB tasks
Cost: $69/month Serpstat + $1.20/month Opus cron

Workflow:

Fetch keywords from my 2 domains
Fetch high-volume keywords in tech/startup niche
Gap analysis (keywords where I DON'T rank)
E-E-A-T analysis (which ones leverage my real experience)
Create NocoDB tasks with priority
Save report to content-strategy/keyword-research/weekly-YYYY-MM-DD.md

The Code (Simplified)

## Pseudo-code for the SEO cron

def weekly_seo_research():
    # 1. Fetch keywords from my domains
    cristiantala_kw = serpstat.get_domain_keywords("cristiantala.com")
    # Returns: 98 keywords

    # 2. Fetch target keywords (tech/startup niche)
    target_keywords = serpstat.get_keywords_top(
        query="startup OR pitch OR investment OR fundraising OR AI",
        country="US",
        limit=500
    )

    # 3. Gap analysis
    my_keywords = set(cristiantala_kw)
    target_set = set(target_keywords)
    gaps = target_set - my_keywords

    # Filter by volume > 100/month AND difficulty < 40
    high_value_gaps = [
        kw for kw in gaps
        if kw['volume'] > 100 and kw['difficulty'] < 40
    ]

    # 4. E-E-A-T Analysis with Opus
    prompt = f"""
    Analyze these keyword gaps for a founder who:
    - Sold a fintech startup for $23M
    - Has 30+ angel investments
    - Is a tech mentor

    Keyword gaps: {high_value_gaps}

    For each keyword:
    1. Relevance (1-10)
    2. E-E-A-T advantage (does their real experience give them an edge?)
    3. Content angle (what unique perspective can they offer?)
    4. Priority (P0/P1/P2/P3)

    Return top 10 opportunities.
    """

    analysis = opus_model.generate(prompt)

    # 5. Create NocoDB tasks
    for opportunity in analysis['top_10']:
        nocodb.create_task({
            "title": f"SEO: Write '{opportunity['keyword']}' post",
            "priority": opportunity['priority'],
            "tags": ["SEO", "Content", "Blog"]
        })

    # 6. Generate report
    report = generate_markdown_report(analysis)
    save_to_file(f"content-strategy/keyword-research/weekly-{today}.md", report)

Sample Report Output (Real)

## SEO Weekly Report — 2026-02-16

### Summary
- **Site keywords ranking:** 98
- **Content gaps found:** 47 high-value keywords
- **Tasks created:** 7 (P1: 3, P2: 4)

### Top Opportunities

#### 1. "pitch deck template" (3.6K/mo, Difficulty 35)
**Angle:** Share actual pitch deck from exit + investor feedback  
**E-E-A-T:** Real exit + 30+ investments = unfair advantage  
**Traffic potential:** 500-800/month  
**Priority:** P1

#### 2. "product market fit" (390/mo, Difficulty 28)
**Angle:** PMF lessons from scaling from 0 to exit  
**E-E-A-T:** Built product from 0 → exit, knows PMF journey firsthand  
**Traffic potential:** 150-250/month  
**Priority:** P1

E-E-A-T: Your Competitive Advantage

E-E-A-T = Experience, Expertise, Authority, Trust (Google's quality guidelines).

My case:

Experience: Sold startup, made 30+ investments
Expertise: 11 years university professor (programming, algorithms)
Authority: Forbes, Bloomberg, major business press coverage
Trust: Real name, verified LinkedIn, public track record

SEO implication:

Do NOT compete on generic keywords ("how to build a startup")
DO compete on keywords where experience = advantage ("negotiate term sheets", "investor due diligence")

Result: I rank higher with fewer backlinks — because Google values E-E-A-T.

Lessons Learned

1. Use Your Best Model for Analysis

I tried a cheaper model for SEO analysis.

Cheap model output: Mechanically lists keywords, misses thematic connections, loses E-E-A-T context.

Powerful model output: "Your exit gives you credibility for 'pitch deck' that competitors lack." Suggests pillar pages, prioritizes by business impact, not just volume.

Verdict: SEO analysis IS strategic thinking. Use your best model.

2. Content Gaps ≠ Keyword Stuffing

190 keyword gaps does NOT mean "write 190 posts."

Better approach:

Group keywords by topic ("pitch deck", "investor deck", "startup presentation" = 1 pillar post)
Create comprehensive content (3,000+ words)
Naturally cover 10-20 related keywords

Result: 10 pillar posts > 100 thin posts.

3. Automate Data, Not Decisions

Serpstat API fetches data → automated.

Which keywords to write about → human decision (AI-assisted).

Anti-pattern: "Just write about top 10 keywords by volume."

Reality: Volume without relevance = wasted effort.

4. E-E-A-T Beats Backlinks (For Personal Brands)

Experiment:

Generic post ("how to make a pitch deck"): 20 backlinks, position #25
Post with real exit story ("pitch deck that closed $23M"): 3 backlinks, position #8

Why: Google detects expertise (exit mentioned, real numbers, unique perspective).

Lesson: For founder content, E-E-A-T > backlinks.

OpenClaw Cron Config

{
  "name": "SEO Weekly Report",
  "schedule": {
    "kind": "cron",
    "expr": "0 1 * * 1",
    "tz": "America/New_York"
  },
  "payload": {
    "kind": "agentTurn",
    "message": "Run weekly SEO research. Fetch keywords via Serpstat API. Identify content gaps, analyze E-E-A-T opportunities, create NocoDB tasks. Save report to content-strategy/keyword-research/weekly-YYYY-MM-DD.md.",
    "model": "opus",
    "timeoutSeconds": 600
  },
  "sessionTarget": "isolated",
  "delivery": {
    "mode": "announce",
    "channel": "telegram"
  }
}

Why Monday 1 AM:

Before the work week starts
SEO tasks ready for Monday morning review
Doesn't interrupt daily workflow

Results (30 Days)

Stats:

Keywords analyzed: 680
Content gaps identified: 190
High-value opportunities: 47
Tasks created: 21 (P1: 7, P2: 10, P3: 4)
Tasks completed: 8 (38% completion in 30 days)

Time saved:

Manual: 3h/week
Automated: 5 min review
Savings: 2h 55min/week = 12 hours/month

ROI Calculation

Cost:

Serpstat: $69/month
Model cron: $1.20/month
Total: $70/month

Value:

Time saved: 12h/month
At $100/hour = $1,200 value
Net ROI: $1,130/month

Payback: Immediate (month 1).

Conclusion: SEO Doesn't Happen By Accident

Manual SEO = "I should do SEO."

Automated SEO = "Here are 7 tasks with clear ROI."

The system converts good intentions into consistent execution.

Result:

12 hours/month saved
Weekly consistency (not "when I have time")
E-E-A-T focus (leverages real competitive advantage)

If you do content marketing and have 50+ posts, automating SEO research pays for itself immediately.

Do you automate SEO research? What tools are you using? Share in the comments.

📝 Originally published in Spanish at cristiantala.com

DEV Community