Local service businesses live and die by search rankings. A sign shop in Boca Raton needs to show up when someone searches "custom signs boca raton" or "vehicle wrap near me." But creating landing pages for every service + location combination manually? That's hundreds of pages.

I built an automated pipeline that mines keywords from Google Autocomplete, stores them in a database, and generates SEO-optimized Next.js pages—then pushes them to GitHub automatically.

The Strategy

1. Mine keywords from Google Autocomplete for service + location combinations
2. Store and dedupe in PostgreSQL
3. Group keywords by service type and location
4. Generate pages when a group has enough keywords
5. Auto-deploy via Git push to Vercel

Mining Keywords from Google Autocomplete

Google Autocomplete is a goldmine. When you type "signs boca raton" it suggests what people actually search for:

async function getAutocompleteSuggestions(query) {
  const url = `http://suggestqueries.google.com/complete/search?client=firefox&q=${encodeURIComponent(query)}`;
  const response = await fetch(url);
  const data = await response.json();
  return data[1] || [];
}

Each query returns 8-10 suggestions. Run 25 queries and get ~200 keywords per batch.

Smart Query Generation

Base queries: Combine service seeds with locations Expand from existing: Take keywords already in your database and generate variations

This creates a feedback loop where your keyword database gets smarter over time.

Page Generation

Group keywords by service + location, then generate pages when you have enough:

for (const [slug, group] of Object.entries(groups)) {
  if (group.keywords.length >= 2) generatePage(slug, group);
}

Results

For a sign shop targeting Palm Beach County:

• 300+ keywords mined from Google Autocomplete
• 25+ SEO pages generated automatically
• Coverage: 8 locations × multiple services
• Time spent: Initial setup only—runs on autopilot

---

Building automated systems that grow your business while you sleep—that's the byldr way.