YouTube Scraping with A-Parser: Where Do Small Channels Succeed?
To test this with actual search data, we used A-Parser to collect 34,070 rows of YouTube search results across 28 topics and 343 search suggestions. Here's what we'll break down:
- Why Search Volume alone is not enough for YouTube;
- How to evaluate competition based on channel sizes in the TOP-10;
- Which three topic groups we identified and where view counts remained low;
- Why a single video appears across 21 search suggestions;
- How to set up niche scraping and creator audits for marketers and media buyers.
If you want to explore the methodology or replicate the experiment yourself, download the keywords, preset, and scraping results (CSV).
1. WHY SEARCH VOLUME ALONE IS NOT ENOUGH FOR YOUTUBE
In traditional SEO, topic research often starts with Search Volume: the higher the monthly search volume, the more attractive a keyword appears.
On YouTube, this metric is insufficient for two main reasons:
1. YouTube combines search and recommendations
Viewers come from their subscriptions, recommendations, search, and Shorts. Our dataset reflects only search results, so it cannot be used to gauge overall viewer interest in a topic.
2. Search Volume does not show how easy it is for a small channel to rank in the top 10
A popular search query does not guarantee a small channel a spot in the TOP-10. If the results are dominated by large channels, high demand alone won't help. However, specific narrow topics can still deliver tens of thousands of views.
Our core practical question: in which topics are small channels already breaking into the top 10 of our scraped search results and generating tens of thousands of views?
2. EXPERIMENT SCALE AND DATASET STRUCTURE
The study began with 28 seed phrases across e-commerce, crypto, fitness, beauty, AI tools, and content creation. For each keyword, A-Parser gathered search suggestions, search results, and metadata on the retrieved videos and channels.
Dataset Overview:
- 28 seed topics;
- 343 "seed keyword + suggestion" pairs;
- 34,070 search result rows;
- 21,537 unique videos (duplicates removed via Video ID v=...);
- 2,366 unique videos in the top 10 of our scraped dataset;
- 4,821 videos (22.38%) found across at least two distinct suggestions.
Out of 34,070 search result rows, we isolated exceptionally successful videos from small channels using four criteria:
- Ranking position: positions 1 through 10 in our export.
- Small channel: 100 to 10,000 subscribers.
- At least 20,000 views: 20,000 views or more.
- Views significantly exceeding audience size: Views/Subscribers = Views / Subscribers ≥ 10.0x.
3. THREE TOPIC GROUPS: WHERE SMALL CHANNELS GET MILLIONS VS. WHERE VIEWS REMAIN LOW
Topics vary significantly by the number of high-performing videos and overall view counts.
Share of breakout videos in each topic:
| Seed Topic | Unique Videos in Top 10 Positions | Breakout Videos | Share of Breakout Videos |
|---|---|---|---|
| iphone unboxing | 108 | 25 | 23.15% |
| weight loss transformation | 84 | 14 | 16.67% |
| home decor aesthetic | 92 | 12 | 13.04% |
| fitness transformation | 93 | 10 | 10.75% |
| neural networks for video | 61 | 6 | 9.84% |
| viral shorts tutorial | 89 | 8 | 8.99% |
| dropshipping 2026 | 43 | 3 | 6.98% |
| best sneakers | 101 | 7 | 6.93% |
| travel vlog | 125 | 5 | 4.00% |
| how to make money on YouTube | 79 | 3 | 3.80% |
| parcel unboxing | 110 | 4 | 3.64% |
| amazon finds | 116 | 4 | 3.45% |
| faceless youtube channel | 69 | 2 | 2.90% |
| youtube automation | 75 | 2 | 2.67% |
| ai video generator | 87 | 2 | 2.30% |
| smartphone review | 133 | 2 | 1.50% |
| affiliate marketing | 109 | 1 | 0.92% |
| crypto exchange review | 136 | 0 | 0.00% |
| shopify dropshipping | 88 | 0 | 0.00% |
| crypto bots arbitrage | 20 | 0 | 0.00% |
The topics fall into three distinct groups.
Group 1: Visual Topics (up to 23.15% breakout videos)
At the top of the table are topics where the video's subject is immediately clear to the viewer:
- Gadget unboxings and aesthetic reviews;
- "Before and after" body transformations;
- Home makeovers and DIY;
- Short demonstrations of AI video creation tools.
Examples from our dataset:
- Tech POP channel (727 subscribers):
- Video: “First iPhone 6 Sold in Perth Dropped by Kid”
- Position: #3 for the suggestion iphone unboxing gone wrong
- Result: 2,721,906 views (Views/Subscribers ratio: 3,744x).
- ipikk channel (374 subscribers):
- Video: “IPHONE 17 Black Aesthetic unboxing (256gb)”
- Position: #4 for the suggestion iphone unboxing aesthetic
- Result: 175,325 views (Views/Subscribers ratio: 468x).
- TikTok Boom channel (729 subscribers):
- Video: “Weight Loss Check Tik Tok Compilation | Motivation”
- Position: #2 for the suggestion weight loss transformation tiktok
- Result: 673,532 views (Views/Subscribers ratio: 923x).
Group 2: Ranking in search is possible – but it doesn't guarantee reach
We expected search results for crypto exchanges, dropshipping, and crypto bots to be dominated almost exclusively by large channels. Surprisingly, videos from small channels made up roughly one in three results in certain topics. But ranking was only half the story: none of those 59 videos reached 20,000 views. In other words, ranking highly in search and actually generating views are two different things. In our study, a video qualified as a breakout only if it reached at least 20,000 views and achieved a views-to-subscribers ratio of at least 10:1.
Group 3: Automated filters lack contextual awareness
Numeric filters alone are not enough. For the query online casino big win, the filter identified matching candidates, but several of the top videos were actually about the virtual casino in GTA 5 Online rather than real-world online casinos.
The filter correctly selected videos based on numeric thresholds, but it did not account for search intent. Therefore, automated filtering is ideal for surfacing potential opportunities, but titles and context must be validated manually.
4. HOW A SINGLE VIDEO APPEARS ACROSS 21 SEARCH SUGGESTIONS
Out of 21,537 unique videos, 4,821 videos (22.38%) appeared across two or more search suggestions.
Several videos appeared for multiple query variants:
- “The 20 Most Profitable Faceless AI Niches Right Now 2026” (Steffen Miro) – appeared for 21 suggestions in the faceless youtube channel cluster.
- “How To Start a Faceless YouTube Channel That Makes Money in 2026” (Joshua Mayo) – also appeared for 21 suggestions.
- “How to Start Shopify Dropshipping in 2026” (Ac Hampton) – appeared for 20 suggestions.
Practical rule for creators:
Do not restrict a video to a single keyword. First gather related search suggestions to understand the broader search topic they collectively represent.
5. FINDING UNDERRATED CREATORS AND BUILDING CONTENT BRIEFS
A-Parser helps build an initial shortlist of micro-creators (1k–50k subscribers) in a target niche for subsequent auditing of pricing, engagement, and audience quality:
- Export search results for a product-focused search suggestion;
- Filter channels with 1k to 50k subscribers;
- Calculate median engagement rate: (Likes + Comments) / Views × 100%;
- Highlight unusually high view-to-subscriber ratios: Views / Subscribers.
How to find ideas and build creative briefs
Successful videos from small channels often answer specific user questions that larger channels overlooked in generic overviews. You can search the data for recurring patterns and translate them into actionable video concepts.
Recurring elements do not prove that a specific tag or title formula directly drove view growth. Instead, they provide data-backed hypotheses that can be tested on relevant videos and manually verified within the niche context.
1. Recurring Tags
In the Tags field of the 127 selected breakout videos, recurring terms described not only the core topic, but also the content format and delivery style:
- aesthetic and roomdecor – visual style in interior design and unboxings;
- unboxingasmr and compilation – video format in tech reviews and fitness transformations;
- makemoneyonline and youtubeautomation – monetization intent in AI tool topics;
- tiktokmemes and motivation – content type and emotional delivery style.
2. Title Modifiers
A generic title like “iPhone 15 Unboxing” competes directly against dozens of identical reviews from major channels. Specifying the exact model, setup, or format helps the video match specific viewing situations.
For instance, the channel Eljohn Reformado Aquino with 951 subscribers amassed 436,704 views and entered the TOP-10 by specifying ASMR format, device setup, and no background music in the title:
Code:
iPhone 15 Unboxing & Setup ASMR (2025) | No Background MusicTranslating data into content briefs:
Instead of vague assignments like "film a product review," leverage the identified formatting patterns:
- Filming format: unboxing with no background music (No Background Music);
- Title structure: exact model + format (Setup & Accessories or Clean ASMR);
- Topic tags: unboxingasmr, aesthetic.
6. NATIVE WORKFLOW IN A-PARSER AND DATA CLEANING
Data is collected in A-Parser through three sequential tasks without writing any code:
During Task B, do not enable URL-only deduplication (Uniq). A single video can appear across dozens of search suggestions. Deduplicating by URL alone will destroy the critical Suggestion ➔ Position ➔ VideoURL mapping.
7. STEP-BY-STEP CHECKLIST FOR MARKETERS AND CONTENT CREATORS
Before committing budget to video production or sponsorships, run a full niche audit:
- Collect 10–30 core keywords for your topic.
- Expand keywords into search suggestions using SE::YouTube::Suggest.
- Scrape search results for suggestions using SE::YouTube, without enabling URL-only deduplication.
- Enrich with video and channel metrics using SE::YouTube::Video.
- Compare channel sizes in the top 10 positions.
- Apply the 4 breakout filters:
- Search position ≤ 10
- 100 ≤ Subscribers ≤ 10,000
- Views ≥ 20,000
- Views / Subscribers ≥ 10.0x
- Manually verify search intent and video relevance, filtering out false positives like GTA.
- Draft content briefs incorporating high-performing formats (ASMR, Aesthetic, Compilation).
8. LIMITATIONS OF THE DATASET AND STUDY
This study has three notable limitations:
- Data type: public YouTube search results lack internal analytics, such as CTR, retention curves, audience geography, or exact traffic sources.
- Point-in-time snapshot: data captures video rankings at the exact moment of scraping without historical tracking.
- Hypotheses: recurring tags and title patterns provide testable insights, but do not guarantee results.
9. KEY FINDINGS OF THE STUDY
Across 34,070 search results, there is no universal niche where every new channel automatically gets millions of views. However, niches vary in search result composition and the number of breakout videos:
- Visual niches: breakout videos from small channels accounted for up to 23.15% of unique videos in the top 10 positions, with individual videos generating between 20,000 and 2.7 million views;
- Commercial topics: videos from small channels made up 33–35% of unique videos in the top 10 positions in select niches, but none of the 59 identified videos hit 20,000 views; the underlying causes fall outside the scope of this study;
- Related suggestions: 4,821 videos (22.38%) appeared across multiple suggestions, with individual videos appearing for up to 21 query variants;
- Limits of automated filtering: metric filters surface high-performing outliers effectively, but examples like GTA 5 highlight why manual intent verification remains essential.
10. HOW A-PARSER REPLACES PORTIONS OF VIDIQ AND SIMILAR TOOLS
Tools like vidIQ help with topic research, video and channel analysis, and competitor monitoring. For analyzing public YouTube search results, many of these workflows can be handled in A-Parser, allowing you to build a custom research setup tailored to your target market rather than relying on standardized reports.
The primary output of A-Parser is not merely a CSV export or a collection of trending videos. A-Parser transforms manual YouTube browsing into a repeatable, structured market research workflow.
Instead of manually entering hundreds of queries, tracking rankings, copying links, and checking subscriber counts, you can preserve the full chain:
Code:
Seed Topic ➔ Search Suggestion ➔ Position ➔ Video ➔ Channel Metadata ➔ Filters & ReportKey research tasks you can bring in-house
- Topic & Keyword Research: collect YouTube search suggestions for seed terms and map the niche's query structure;
- Competition Analysis: analyze search rankings, channel sizes, views, and Views/Subscribers ratios;
- Creator & Format Discovery: discover small channels, top-performing videos, recurring tags, and title modifiers;
- YouTube Search Results Monitoring: run periodic data pulls to compare snapshots and track how rankings and TOP-10 composition change over time;
- Custom Metrics & Reporting: apply tailored thresholds, filters, and export formats aligned with your product, niche, or media buying hypotheses.
You retain full control over key terms, collection schedules, filter logic, and export formats – giving you a YouTube research system tailored to your niche, product, or media buying strategy.
For analyzing public YouTube search results, this framework can replace equivalent features in vidIQ or similar tools while using your own filters, thresholds, and metrics.
Product Stack for YouTube Research:
- A-Parser Pro / Enterprise
A tool for collecting search suggestions (SE::YouTube::Suggest), search results (SE::YouTube), and video/channel metadata (SE::YouTube::Video). Perpetual license with no monthly subscription. - A-Parser Unlimited Proxies
A solution for recurring, high-volume scraping. Pricing is based on thread count without bandwidth caps, making budget forecasting straightforward. - A-Parser Premium Residential Proxies
A pool of residential IPs for geo-specific checks. This experiment did not isolate proxy type or geographic targeting, so we did not evaluate their impact on the results.
Useful Links & Resources:
Last edited: