Decodo Scraper API · target=google_maps + custom DOM parse, anchored on Places-API-geocoded outlet lat/lng. $0.001 per pin, 5-8 s, returns 1-20 rank depth with 13 per-place fields (rating, reviews, category, address, hours). 5× the brand-visibility coverage of the Search local-pack 3-pack at the same cost.
Live Decodo Scraper API runs on real outlets. Each card has a thumbnail of the area (Google Static Maps) and links to the interactive heatmap. Hover any pin on the heatmaps to see the top-10 ranked competitors with ratings.
Indiranagar · Method C (Maps top-20) · Bangalore
5×5 · 1 km · 25 pins
Toit dominates craft-beer queries across Bangalore. Hover pins to see the full Bier Library/BLR Brewing/Indian Biere House landscape Toit is competing in.
| keyword | coverage | SoLV | AGR |
|---|---|---|---|
| best brewery bangalore | 88% | 56% | 2.95 |
| craft beer bangalore | 96% | 96% | 1.00 |
Indiranagar (extended) · Bangalore
11×11 · 1.5 km · 121 pins
11×11 1.5 km — Toit's brand catchment across the city for craft-beer queries.
| keyword | coverage | SoLV | AGR |
|---|---|---|---|
| best brewery bangalore | 98% | 3% | 8.83 |
| craft beer bangalore | 43% | 42% | 2.17 |
Indiranagar (flagship brewpub) · Bangalore
7×7 · 1 km · 49 pins
Even branded queries return rank 2-4 here — useful demo of yellow/green mixing across pins.
| keyword | coverage | SoLV | AGR |
|---|---|---|---|
| toit indiranagar | 100% | 0% | 4.00 |
| best brewery bangalore | 100% | 18% | 7.82 |
| craft beer indiranagar | 100% | 100% | 2.29 |
Sangam Courtyard · Method C (Maps top-20, geocoded) · Delhi NCR
5×5 · 0.5 km · 25 pins
5×5 / 0.5 km. Outlet coords resolved via Google Places API (place_id ChIJbZCrwQQdDTkRC7ARXvyMSsY). Previous coord guess was 1.4 km off — corrected version shows 72% SoLV for category queries.
| keyword | coverage | SoLV | AGR |
|---|---|---|---|
| best indian restaurant | 100% | 72% | 3.00 |
| best north indian restaurant | 88% | 0% | 5.23 |
Vasant Kunj · Method C (Maps top-20) · Delhi NCR
5×5 · 1 km · 25 pins
5×5 grid via Method C. Hover any pin: side panel shows top 10 actual competitors at that lat/lng — Dilli Kanteen, Mala-Akbari, AnnaMaya, etc. Anardana row is highlighted blue when present.
| keyword | coverage | SoLV | AGR |
|---|---|---|---|
| best indian restaurant | 88% | 8% | 6.55 |
| best north indian restaurant | 68% | 4% | 6.41 |
Sangam Courtyard, RK Puram · Delhi NCR
5×5 · 0.3 km · 25 pins
Tight 5×5 at 0.3 km — the outlet owns its block on category keywords, not just brand.
| keyword | coverage | SoLV | AGR |
|---|---|---|---|
| anardana | 100% | 100% | 1.00 |
| best indian restaurant | 100% | 100% | 1.88 |
| fine dining | 0% | 0% | — |
Ambience Mall, Vasant Kunj · Delhi NCR
11×11 · 1.5 km · 121 pins
11×11 grid spanning ~15 km — Anardana's catchment ring vs the rest of South Delhi.
| keyword | coverage | SoLV | AGR |
|---|---|---|---|
| best indian restaurant | 21% | 21% | 2.04 |
| best north indian restaurant | 12% | 12% | 2.29 |
Ambience Mall, Vasant Kunj · Delhi NCR
7×7 · 1 km · 49 pins
Branded queries dominate; category queries show clear geographic drop-off past ~2 km.
| keyword | coverage | SoLV | AGR |
|---|---|---|---|
| anardana | 100% | 100% | 1.00 |
| best indian restaurant | 39% | 39% | 2.05 |
| best north indian restaurant | 18% | 18% | 2.78 |
Ambience Mall, Vasant Kunj · Delhi NCR
9×9 · 1 km · 81 pins
Same outlet, wider grid — the circular fade-out around the mall becomes obvious.
| keyword | coverage | SoLV | AGR |
|---|---|---|---|
| anardana | 100% | 100% | 1.00 |
| best indian restaurant | 40% | 40% | 1.72 |
Jama Masjid, Old Delhi · Delhi NCR
9×9 · 1.2 km · 81 pins
Iconic single-outlet brand in dense urban core — 9×9 1.2 km shows how it owns Old Delhi.
| keyword | coverage | SoLV | AGR |
|---|---|---|---|
| karims | 100% | 100% | 1.00 |
| best mughlai food old delhi | 0% | 0% | — |
Bandra coords (no outlet) · Mumbai
7×7 · 1 km · 49 pins
Negative reference — the brand doesn't operate in Mumbai. Honest all-grey heatmap.
| keyword | coverage | SoLV | AGR |
|---|---|---|---|
| best italian restaurant | 0% | 0% | — |
| italian food | 0% | 0% | — |
| best pasta | 0% | 0% | — |
Pali Hill, Bandra · Mumbai
5×5 · 0.5 km · 25 pins
Bakery chain — branded query is uniform #1, generic 'best bakery' doesn't surface them.
| keyword | coverage | SoLV | AGR |
|---|---|---|---|
| theobroma bandra | 100% | 100% | 1.00 |
| best bakery mumbai | 0% | 0% | — |
| brownies bandra | 0% | 0% | — |
Vasant Kunj · same 25 pins, both methods · Delhi NCR
25 pins · 2 methods overlaid
5.5× brand visibility (C vs A)
| method | coverage | depth |
|---|---|---|
| A · Search local-pack | 16% (4/25) | top 3 |
| C · Maps top-20 | 88% (22/25) | top 20 |
Indiranagar · same 25 pins, both methods · Bangalore
25 pins · 2 methods overlaid
0.9× brand visibility (C vs A)
| method | coverage | depth |
|---|---|---|
| A · Search local-pack | 100% (25/25) | top 3 |
| C · Maps top-20 | 88% (22/25) | top 20 |
Four delivery paths for geo-anchored Google ranking data — all tested empirically. Same lat/lng went into every method (verified across a 7-variant config probe).
| method | A · Search local-pack | B · Camoufox + IN proxy | C · google_maps + parse |
|---|---|---|---|
| endpoint | Scraper API google_search · parse=true | Camoufox over Decodo ISP proxy → Maps DOM | Scraper API google_maps · custom regex parse |
| cost / query | $0.001 | ~$0 (proxy bandwidth only) | $0.001 |
| speed / query | 2-4 s | 15-30 s | 5-8 s |
| rank depth | 1-3 (local-pack only) | 1-20 | 1-20 |
| per-place fields | name + position | name (DOM scrape) | 13 fields incl. rating, reviews, category, address, hours |
| infrastructure | none | headless worker pool + proxy mgmt | none |
| geo-anchoring | server-spoofed to lat/lng/radius | real proxy IP + Maps viewport URL | server-spoofed to lat/lng/radius |
| elapsed (9 pins) | 5s | n/a | 24s |
| matched in (9 pins) | 7/9 | n/a | 5/9 |
| verdict | misses 80% of brands ranking 4-20 | backup / verification only | ← V1 primary |
5× coverage finding (Anardana Vasant Kunj, 25 pins): Method A found the brand at 4/25 pins (16%); Method C found it at 21/25 (84%) — anywhere in top-20. Same cost, same lat/lng, same Decodo plan. Method A is structurally blind to ranks 4-20 because the Search local-pack only shows 3.
Why same lat/lng disagreed earlier: we were comparing Search rad: 1000m against Maps viewport zoom=14z — different geo-scope conventions, not different rankings.
A 7-variant controlled probe showed that locale, domain, and radius barely affect Scraper API output; lat/lng is the only knob that moves the local pack.
Subagent verified pricing live on 2026-05-27 across 12 providers. Full brief in docs/google-data-sources-brief.md.
| provider | $/1M queries | depth | lat/lng? | verdict |
|---|---|---|---|---|
| DIY · Camoufox + Webshare/Decodo proxy | $300-800 | 100 | via headless URL | brittle, ops-heavy |
| Serper.dev (est.) | $300 | 100 | yes | fastest 1-2s; "scan now" demos |
| DataForSEO Standard (Search) | $600 | 100 | yes | V1.5 — Search-side cost cut |
| Decodo Scraper API (Search + Maps) | $1,000 | 100 | yes | V1 primary (what we ship) |
| ValueSERP top tier | $1,000 | 100 | yes | Tier 2 fallback |
| Oxylabs SERP | $1,000-1,350 | 100 | yes | enterprise alt |
| Bright Data SERP (est.) | $1,500 | 100 | yes | premium fallback |
| Apify Google Maps | $2,100 | 120+ | yes | too slow daily |
| SerpApi | $9,000 | 100 | yes | 10× our cost — hard pass |
| ScraperAPI | ~$12,500 | DIY | DIY | hard pass |
| ScrapingBee | $14,700 | DIY | DIY | hard pass |
| Google Places API | varies | n/a | n/a | not a ranking API · used for outlet geocoding only |
| Ahrefs API | per-unit | their index | city only | keyword research only |
Each sprint compounds raw-data depth on the same $0.001 Decodo query — no new spend, just better extraction + compute.
Two-surface parser (DOM regex + "Suggest an edit" data blob) emits 13 fields/place: name, KG_ID, ChIJ_ID, rating, review_count, category, full address, hours, phone, website, operating status, price tier, photo, review snippet.
Review-Weighted Visibility (RWV = Σ rank_weight × log(review_count)), competitor density per pin, category-mix donut, open-now coverage, rating-gap analysis. All from Sprint-1 data — zero new queries.
est. 1 day · differentiates from BrightLocal/Local Falcon
Postgres 16 + Timescale-compatible schema (hypertable at 10M rows). Leaflet TimeDimension for weekly playback, custom Canvas delta-arrows. Risk flagged: alert fatigue — needs per-customer threshold tuning before rollout.
Pulled live via Ahrefs MCP to validate which keywords Anardana actually owns organically (India, May 2026).
Every one of Anardana's top organic keywords is branded. That's why the category-keyword layers in our heatmaps (e.g. "best indian restaurant") reveal real geographic variation — Anardana doesn't own those queries at scale.
| keyword | volume / mo (IN) | best pos | traffic / mo | flags |
|---|---|---|---|---|
| anardana restaurant | 3,800 | 1 | 1,211 | branded |
| anardana chandigarh | 5,800 | 2 | 846 | branded · local |
| anardana ranchi | 2,300 | 4 | 213 | branded · local |
| anardana gurgaon | 2,000 | 4 | 191 | branded · local |
| anardana restaurant delhi | 800 | 2 | 138 | branded · local |
| anardana sangam courtyard | 1,200 | 4 | 108 | branded · local |
| anardana karkardooma | 450 | 4 | 41 | branded · local |
| anardana menu | 100 | 1 | 38 | branded |