Starvia
10,000+
First-Month Google Impressions
33%
MENA Arabic AI Mention Rate
25%
LatAm Spanish AI Mention Rate
01
Client Background
Starvia is a Chinese auto dealer serving South America in Spanish and MENA in Arabic, covering both new and used vehicle export. As a dealer between OEMs and end buyers, Starvia faced a built-in dilemma: specs and model data already live on OEM sites, so dealer pages lacked differentiated value and neither search engines nor AI assistants had a strong reason to surface them. At kickoff, monthly Google impressions were just 14, and AI-answer presence was near zero.
02
Goals & Challenges
Project Goals
The client needed to build, from zero, a multilingual independent site that overseas buyers could find and AI assistants could cite. The project also needed a CRM and traffic-source tracking so every lead, content asset, and market signal could be attributed. Target markets included LatAm Spanish and MENA Arabic, requiring minority-language content, localized topic selection, and Arabic RTL support.
Core Challenges
If auto dealer content only repeats model specs, it cannot beat OEM sites. LatAm Spanish and MENA Arabic are also high-barrier minority-language markets where English content advantages do not transfer directly. The project had to build organic search visibility, content assets, and AI visibility from a cold start while finding the local value that dealer pages could uniquely provide.
03
Strategy & Execution
Core Insight
Writing what the car is will never beat the OEM site. The real differentiation is writing how the car is used locally. Organizing content around local language, roads, climate, commuting, imports, and dealer decision scenarios gives dealer pages local value that OEM sites cannot provide.
GEO Execution
Analyzed how Spanish-speaking LatAm and Arabic-speaking MENA users actually ask vehicle and import questions, then reorganized content around local language and local scenarios. Topics covered Middle-East desert and city-road performance, long-haul reliability, daily commuting, local model comparisons, and Chinese SUVs suited for regional dealers, helping Starvia enter AI recommendation contexts in minority-language markets.
SEO Execution
Starting from the client's logo, visual assets, and vehicle inventory, we designed and engineered the site, built a four-language architecture (ZH / EN / ES / AR, including Arabic RTL), and deployed a CRM to connect the full traffic-source chain. On the content side, the project produced 20 scenario blogs in each of four languages and a multilingual FAQ library with 500 images.
Execution Timeline
The first live month delivered the site, CRM, and multilingual content foundation from zero. Google impressions grew from 14 to 10,062 in that month; by partial July, average position had already improved from 14.0 to 10.1, while clicks rose from 1 to 81. The project is now in its second month of deeper optimization.
04
Project Results
Cumulative impressions after launch
Daily cumulative, counted from the day it went live
Vertical axis: cumulative impressions, unlabelled; 84 days of daily GSC data from 2026-05-27
How to read: the vertical axis is relative magnitude with absolute values withheld; the curve expresses shape and timing only, while the magnitudes are listed in the results table below. The curve stops at its peak — later changes in pacing reflect client-side budget and scheduling, not a decline in effectiveness.
| Metric | Change |
|---|---|
| Google Impressions (GSC) | From 14 to 10,000+ in month one; average position keeps climbing |
| MENA Arabic AI Mention Rate | From 0 to 33% |
| LatAm Spanish AI Mention Rate | From 0 to 25% |
| Markets Covered | LatAm (Peru / Bolivia / Argentina / Colombia / Chile) + Spain + MENA (Algeria, etc.) |
| Content Assets | 4-language scenario blogs (20 each) + multilingual FAQ library with 500 images |
| Delivery Scope | Site + CRM + multilingual GEO/SEO content, one-stop delivery |
Data source: Google Search Console / brand AI-visibility monitoring
More Case Studies
AI Database
Zilliz & Milvus
A vector database brand going global — overtook industry leader Pinecone in AI mention rate and became #1 in organic search traffic.
Cloud Platform
Sealos
An AI-native cloud platform going global — achieved 10x Google search click growth in 6 months with 50%+ AI mention rate.
Academic Search
WisPaper
An AI academic search engine going global — Google impressions grew tens of times over, AI recommendations reached mainstream status.