Local Pack, Map Results, and the Local SEO World

For any query with local intent, "pizza near me," "dentist Chicago," "stores near 90210", Google injects the Local Pack: 3 nearby businesses with a map. Behind it is Google Maps, Google Business Profile (formerly Google My Business / GMB), and a parallel ranking system distinct from organic SEO.

For a scraper, this is one of the most valuable SERP blocks. It carries phone numbers, addresses, hours, ratings, directly actionable data.

What the Local Pack contains

Visually: three business cards above the regular organic results, each with name, rating, distance, brief metadata, and a map showing their locations.

Sample SERP-API shape:

{
  "local_results": {
  "places": [
  {
  "position": 1,
  "title": "Mario's Pizza",
  "place_id": "ChIJ...",
  "rating": 4.6,
  "reviews": 1234,
  "price": "$$",
  "type": "Pizza restaurant",
  "address": "123 Main St, Chicago, IL 60601",
  "phone": "(312) 555-0100",
  "website": "https://...",
  "hours": "Open ⋅ Closes 11 PM",
  "service_options": {"dine_in": true, "takeout": true, "delivery": false},
  "gps_coordinates": {"latitude": 41.881832, "longitude": -87.623177},
  "thumbnail": "..."
  }...
  ],
  "map_url": "https://www.google.com/maps/search/..."
  }
}

Key fields:

• place_id, Google's permanent ID for the business. Stable across queries.
• gps_coordinates, useful for plotting / geo-analysis.
• service_options, dine-in, takeout, delivery, curbside.
• hours, current-day text status. For full hour schedules, hit the Place Details API or scrape the Place page.

Local Pack vs Map Results

Two distinct things:

• Local Pack = the 3 businesses on the main SERP. Cheaper to scrape; surfaced for every local query.
• Map Results = the full Google Maps listing when you click "More places" or visit google.com/maps?q=.... Returns 20–60 places, sorted by relevance + distance + rating.

Many SERP-APIs support both. Use Local Pack for SEO/SERP analysis ("does my business rank for X near Y?"); use Map Results for full-market lead generation ("all pizza places within 5 miles of 60601").

Catalog108's local-pack mock

practice.scrapingcentral.com/search?q=stores+near+me returns a SERP with a local-pack-style block. The data comes from /api/locations:

import requests

r = requests.get("https://practice.scrapingcentral.com/api/locations", params={"q": "stores near me"})
locations = r.json()

for loc in locations.get("places", []):
  print(f"{loc['name']} ({loc['rating']}★), {loc['address']}")

Catalog108 ships its locations as JSON for easy practice. Real Google requires a SERP-API.

The Google Maps XHR (for the local pack itself)

When you visit Google Maps in a browser, the page makes XHR calls to internal endpoints, extremely complex protobuf-encoded responses. Scraping it directly is hard:

• Endpoints like https://www.google.com/maps/preview/place returning protobuf.
• Token systems and per-session cookies.
• Heavy anti-bot.

SERP-APIs do this work for you. Lesson 3.30 covers why direct Maps scraping is impractical.

A local-lead-generation use case

Suppose you want all dentists within 10 miles of every ZIP code in a state:

def fetch_dentists_in_zip(zipcode, api_key):
  r = requests.get("https://serp-api.example.com/search", params={
  "q": "dentist",
  "engine": "google_maps",
  "location": zipcode,
  "type": "search",
  "api_key": api_key,
  })
  return r.json().get("local_results", [])

all_dentists = []
for zipcode in load_state_zipcodes("CA"):
  all_dentists.extend(fetch_dentists_in_zip(zipcode, api_key))

# Dedupe by place_id (the same dentist near multiple ZIPs)
seen = set()
deduped = []
for d in all_dentists:
  pid = d.get("place_id")
  if pid and pid not in seen:
  seen.add(pid)
  deduped.append(d)

print(f"Found {len(deduped)} unique dentists in CA.")

PHP equivalent: same shape with Guzzle + array_unique on place_id.

place_id is critical for de-dup, the same business surfaces for multiple nearby ZIPs.

Local SEO context

The local pack ranks by a different signal blend than organic:

• Proximity, how close to the searcher's location.
• Relevance, does the GMB category match the query.
• Prominence, review count, review velocity, citations across the web.

If you're tracking local rankings, your scraper must:

• Specify a precise location (latitude/longitude, not just "Chicago").
• Track over time (proximity changes with map rendering edge cases).
• Capture both "rank 1 in local pack" and "rank N in map results", they're different KPIs.

Parsing hours

Hours from a single SERP call is just "Closes 11 PM today." For the full weekly schedule, you need:

• Google Place Details API (paid, requires Google Cloud).
• A SERP-API's place details endpoint (most SERP-APIs offer one).
• Scraping the place's Google Maps page (hard, see above).

For most use cases, the SERP-API's place details call is the right balance.

A local-pack parser

def parse_local_pack(data: dict) -> list[dict]:
  return [
  {
  "place_id": p.get("place_id"),
  "name": p.get("title"),
  "rating": p.get("rating"),
  "review_count": p.get("reviews"),
  "address": p.get("address"),
  "phone": p.get("phone"),
  "website": p.get("website"),
  "lat": p.get("gps_coordinates", {}).get("latitude"),
  "lng": p.get("gps_coordinates", {}).get("longitude"),
  "rank": p.get("position"),
  "category": p.get("type"),
  }
  for p in (data.get("local_results", {}) or {}).get("places", [])
  ]

Hands-on lab

Open practice.scrapingcentral.com/search?q=stores+near+me. Note the local-pack-style block. Fetch /api/locations?q=stores+near+me and parse each location into a dict with name, address, phone, rating, lat/lng. Then think about: if you wanted dentists across an entire state, how would you de-dup across overlapping searches? (Answer: by place_id.)