YouTube Shorts Public Search Benchmark 2026

Most Shorts “benchmarks” become misleading when they skip the data boundary. A public search page can show a Short, title, public view text, thumbnail, and source URL. It cannot show another channel’s retention graph, chose-to-view rate, revenue, RPM, audience mix, or conversion data.

This asset is intentionally narrow. It documents a reproducible public sample that creators and researchers can inspect, challenge, or repeat.

Dataset snapshot

Download the public appendices:

• CSV observations
• Summary JSON

Sampling method

We fetched these three public YouTube search URLs from the ContHunt SEO VPS using a desktop browser user agent:

1. https://www.youtube.com/results?search_query=youtube+shorts+content+ideas&sp=EgIQCQ%253D%253D
2. https://www.youtube.com/results?search_query=youtube+shorts+editing+tips&sp=EgIQCQ%253D%253D
3. https://www.youtube.com/results?search_query=youtube+shorts+analytics&sp=EgIQCQ%253D%253D

The sp value came from YouTube’s own public search filter labeled “Shorts” in the fetched result page. We parsed the embedded ytInitialData response and extracted only shortsLockupViewModel entries.

Observed fields

Each raw row preserves:

• collection timestamp;
• source query;
• source search URL;
• rank within the fetched response;
• YouTube video ID;
• Shorts URL;
• visible title text;
• public view text exactly as YouTube served it;
• thumbnail width and height when present;
• YouTube page type.

We kept raw rows and deduplicated summary counts separately. The CSV keeps all 79 observations because duplicates across queries are part of the public search result experience.

What this sample can say

In the 74 unique Shorts, simple title-string checks found:

These counts describe this sample only. They are not a formula for distribution and they do not prove why any Short received views.

Sample rows

The CSV preserves the exact localized public view text returned to the server. The table above trims some symbols for readability and links back to the original Shorts.

Exclusions

This sample does not include:

• YouTube Studio retention, shown-in-feed, chose-to-view, or audience metrics;
• private channel analytics;
• RPM, revenue, or monetization rates;
• paid promotion data;
• customer data;
• backlink, ranking, or authority claims;
• normalized numeric view counts.

We did not normalize localized view text into integers because the goal was reproducibility from public fields, not a scraped performance leaderboard.

How to reproduce it

1. Open the three source URLs on the same date or a new declared date.
2. Keep the Shorts type filter selected.
3. Save the raw public HTML response.
4. Parse only shortsLockupViewModel entries from ytInitialData.
5. Keep the source query, rank, video ID, Shorts URL, title, public view text, thumbnail dimensions, and page type.
6. Disclose locale, date, and any blocked or missing fields.

If you repeat the collection later, expect the rows to change. YouTube search results are personalized, localized, and time-sensitive.

How ContHunt uses this kind of sample

Use a sample like this to build research questions, not universal rules. For example:

• Which visible titles rely on tutorial promises?
• Which titles use a number or list?
• Which titles make a money, growth, viral, subscriber, or view claim?
• Which examples are worth opening for hook, pacing, caption, and visual structure review?

Then test one variable on your own channel and measure it in YouTube Studio. Public creative research and private channel analytics answer different questions.

Bottom line

This benchmark is linkable because it is small, dated, source-labeled, and honest about what it cannot know. The useful takeaway is not that a title pattern guarantees views. The useful takeaway is the method: collect public Shorts with source URLs, write down only observable fields, and keep unsupported analytics claims out of the conclusion.