Road Noise · Report 01 · April 2026
42 reading now

The Top 50+ Car Brands on social video.

Most mentioned car brands across Instagram and TikTok in 9,970 short-form videos, six languages, frame-level. Plus how the conversation actually splits — performance, lifestyle, EVs, the Chinese wave.

Analysis by
Stéphane Bonnegent
SB
Stéphane Bonnegent
Founder, Stories Engine

Social media and audience analysis expert. Helps companies identify their highest-potential segments and build content strategies that convert.

Section 01 · The Ranking

Ranking by brand.

Every brand mentioned in three or more videos across the dataset, ranked by total mentions, views, interactions, or share of mentions that happened only in spoken audio. Each video counts a brand once even if mentioned multiple times.

Sort by
Filter by language
# Brand Country Mentions Views Interactions Shadow reach %
About this dataset
Section 02 · By model

Ranking by model.

Same dataset, model-level. The 911 isn't surprising. The Civic at #2 by mentions might be. Sort and filter the same way as brand-level. Every model mentioned in three or more videos.

Sort by
Filter by language
# Model Country Mentions Views Interactions Shadow reach %
Section 03 · Shadow Reach

What people actually say about cars.

Most listening tools index the caption — what creators type. Half of brand mentions never appear there. They live only in the spoken audio. That's shadow reach. Below: the conversation broken into six topics, the way people on social actually talk about cars.

51%
Average shadow reach
Of all brand mentions across the dataset, this share happened only in spoken audio. Caption-only listening would miss them.
9,970
Videos analyzed
Frame-level visual analysis paired with full audio transcription, across six languages, on Instagram and TikTok.
79
Brands tracked
Plus 178 distinct car models, with disambiguation rules for ambiguous tokens like "Smart" or "VW".
51%
of brand mentions live only in spoken audio. The other half of the conversation is invisible to caption-based tools.

What people talk about, by topic.

Share of brand-mention videos that touched each topic. A single video can cover multiple topics — performance and lifestyle often overlap. Pick a brand to see how its conversation differs from the average.

Section 04 · Electric

What people actually look at in an EV.

Every legacy automaker is trying to sell EVs into a saturated market. Below: what people on Instagram and TikTok actually evaluate when an electric car appears on screen — across EV-relevant videos.

The buying criteria, ranked.

Share of EV-relevant videos that mention each aspect. Range still leads. Tech and screens are catching up. Charging speed and battery life — the things engineers care about most — get a fraction of the airtime.

Top EV brands by mentions.

What people are saying.

A handful of EV-tagged videos from the dataset, ranked by views. Translations are added below non-English captions.

Section 05 · Big shift

The Chinese wave.

European OEMs are watching the rise of Chinese auto brands with concern. Chinese brands appeared in the dataset, named in videos. Below: which ones, what people actually say about them, and a sample of the conversation.

Chinese brands present
From mass-market BYD and Geely to enthusiast-tier Zeekr, premium Hongqi, and the new wave: Xiaomi, Jaecoo, Omoda, Voyah.
Videos in the conversation
Across all six languages, where a Chinese auto brand was named or where a "rising Chinese brand" narrative was used.
In Spanish-language
Spanish accounts for the largest share of Chinese-brand mentions in the dataset — a leading indicator for Latin American adoption.

Chinese brands in the dataset.

What people actually say about Chinese cars.

Share of Chinese-brand videos that surface each angle. Range and performance lead. Skepticism and "beats Western brands" are real but small. The conversation is more constructive than European brand teams might expect.

Voices.

A sample of the most-watched Chinese-brand videos in the dataset. Non-English captions are translated to English in italic below.

Section 06 · How this was built

How we hear it.

This report wasn't built with hashtag tracking or caption-only listening. It was built with frame-level video search — vision models that watch what's on screen, audio transcription that captures what's said, across every public video on Instagram and TikTok.

Caption listening misses the half that matters.

Most social listening tools were built when video was a small share of social. They rely on hashtags, post text, and bio metadata. That worked when posts were text and images — it doesn't work when 70% of attention is on Reels and TikTok.

When we tracked car brands across both layers — caption and spoken audio — we found that 51% of mentions happened only in audio. Half the conversation is invisible to anything that doesn't transcribe and search the audio.

  • Frame-level vision indexing — what's actually on screen
  • Spoken-word transcription — what's actually said
  • Six languages, full-fidelity, every public video

What you can ask Oriane.

  • "Find every video where someone is driving a {your brand} and saying anything about reliability."
  • "Pull every Reel where a competitor's car appears for more than 3 seconds, ranked by views."
  • "Surface conversations about charging anxiety in Spanish-language videos in the past 30 days."
  • "Show me which creators are organically driving the conversation about your category."
See Oriane in action →
Report by
Stéphane Bonnegent
SB
Stéphane Bonnegent
Founder, Stories Engine

Social media and audience analysis expert. Helps companies identify their highest-potential segments and build content strategies that convert.

See it on your category

51% of the car conversation lives in audio. Yours probably does too.

This report covered cars. The same hidden layer exists in beauty, fashion, luxury, tech, food & beverage — anywhere creators talk to camera. Book a 30-minute demo and we'll run a live shadow-reach pull on your brand and three competitors.

Book a demo with Oriane
Methodology. 9,970 short-form videos collected via Oriane's frame-level video search, on Instagram and TikTok, across six languages (English, French, German, Italian, Spanish, Dutch). Each video must visually contain a car (validated by AI Vision) and contain a synonym for "car" in the spoken audio. Brand and model detection runs on both caption text and spoken-word transcripts, with disambiguation rules for ambiguous tokens. Topic detection uses six conversation categories aligned to the way people actually talk about cars on social. Counts are unique per video — a brand mentioned multiple times in one video counts as one mention. Engagement Rate is per video views. Shadow Reach is the share of brand-mention videos in which the brand was named only in the spoken audio, never in the caption.