
YouTube Tests Ask YouTube AI Answers at the Top of Regular Search Results
YouTube is testing Ask YouTube AI answers at the top of standard search results for how-to and open-ended queries. Here’s who sees it and what it means for video SEO.

Google Search Console's new Platform properties went live for one agency's YouTube channel on July 7. On the first day alone it recorded 18,233 impressions from Google Search. Eleven days in, the running total passed 200,000.
Clicks across those 11 days: 87.
Extrapolate that run rate across a year and it comes to 6.7 million impressions and roughly 2,900 clicks never previously visible.
That gap is the single most useful thing the new report shows. Google already considers the channel eligible to appear for roughly 18,500 searches a day.
If your channel has been active for a year or more, you almost certainly have the same shape of data waiting — real query demand your social content is already surfacing for, invisible until now.
Different populations, different intents — and the second was a black box until July. Instagram and TikTok analytics have the same blind spot, and you can integrate those too.
Nobody briefs a YouTube video to rank on Google. Creators optimise for views, watch time and subscribers while search sits in a separate workstream.
On July 7, Google's Moshe Samet announced platform properties in Search Console. Connect a YouTube channel, Instagram profile or TikTok account and you see the Google Search queries where that platform's content appears, along with impressions, clicks and positions. The feature itself is covered in Search Console Platform Properties Go Global.
On its own, that is reporting. What makes it matter is what Google is doing at the same time.
AI Mode is increasingly multimodal. It reads images and video directly, fans out queries from them, and assembles answers from multiple formats as first-class results, not decoration beneath the blue links.
Which means one query can give a brand several placements at once. A brand publishing only text competes for one slot. A brand mapped across formats can hold several on the same results page.
The export showed 149 distinct phrasings of "enterprise seo" totalling 31,555 impressions in 11 days. The exact phrase sat at position 10.7 — one nudge off page one.
No single variant would survive a keyword shortlist. Together, they were one of the channel's biggest demand pools.
Videos never briefed as search plays were sitting on page one, outperforming the agency's own website on the same queries.
Eighty-three separate videos earned web search impressions in the first 11 days. The top 1,000 queries produced 149,220 impressions and 10 clicks. Wide and shallow — exactly what "eligible but not packaged" looks like at scale.
| Query | Video position | Video impressions | Website impressions |
|---|---|---|---|
| ai search | 7.6 | 7,026 | 1 |
| what is a sitemap | 4.4 | 3,863 | 0 |
| enterprise seo | 10.7 | 3,226 | 1,470 |
| seo specialist | 28 | 4,953 | 3,622 |
| keyword research | 20 | 3,224 | 36 |
The videos are not duplicating the site's visibility. They are earning visibility the site never had.
The channel's own history provides an instructive contrast.
Early videos were unapologetically low production. Filmed at home with no real setup, showing the author on camera walking through tasks like editing an htaccess file.
The theory was simple: find top-of-funnel queries and make a video that answers them. "How to change your slug on WordPress" got exactly that, no frills. Videos were briefed the way pages were briefed back in 2021.
After a break, the channel relaunched in 2026 using the opposite playbook.
Study similar-sized channels, find the outlier videos (the ones doing five or ten times a channel's normal numbers), and make a better version. Like the classic Skyscraper method, but for video.
In that playbook, packaging comes first:
The meta also runs on repetition, or "reps." Consistent uploads for months, gaining single-digit subscribers per video, until the algorithm figured out who the audience was and distribution flipped. The newest upload reached 2,300 views in nine days, several multiples above videos from three months earlier.
Broad top-of-funnel topics win that game.
Platform property data and 12-month analytics confirmed it.
The top content table is still dominated by the old search-first videos.
The clicked videos report tells the same story at smaller scale. "What is AI Search?" leads with eight clicks and "Keywords No Longer Work the Way You Think in 2026" earned five — both relaunch videos. Between them sit the old guard: the January 2023 htaccess how-to with six, the 2022 WordPress slug walkthrough with five, and "SEO Specialist Skills" with three.
Even the channel page itself earned four clicks nobody planned for. One came from the query ".htaccess wordpress."
Looking back at the home-filmed video, and cringing at the quality, it's still converting searches years later.
The feed rewards packaging, breadth, and consistency, and pays out fast. The SERP rewards matching a query and staying relevant to it, and pays out for years.
A channel that only plays the YouTube meta leaves the search game unplayed. The social topical map is how you play both on purpose.
SEO teams have used topical maps for years. The short version: stop chasing keywords one at a time. Map every subtopic and question a subject contains, cover the map completely, and the engine starts treating you as the authority on the entire topic.
Social teams, meanwhile, have run content calendars by platform. The two have never shared an evidence base.
The social topical map applies the topical authority and coverage playbook SEO teams have refined for years to social and video content, built from the platform property data Google just handed us.
One map drives content across YouTube, Instagram, TikTok, X and LinkedIn.
In Search Console, add each platform account you control as a property and verify it.
Export the query data once you have collected a few weeks of it.
First, group phrasings into clusters before judging anything. The "enterprise seo" cluster looked like nothing as individual queries and 31,555 impressions as a group.
Then triage every cluster into one of three moves:
Combine three inputs: the platform queries from Step 2, standard keyword research, and Google Trends for direction.
Organise the output by topic cluster, not by platform. Each cluster gets its demand evidence, dominant intent and the formats that fit it.
This is the structural change. The seed of planning becomes the topic, and platforms become distribution channels for it.
Match format to intent within each cluster:
One cluster produces several assets across platforms, all pulling toward the same demand instead of five teams guessing separately.
The fan-out gives you the split: the topic is the blog pillar, a facet is the long-form video, and a sub-facet is the short, cut from the video you already made.
Using creatine as an example:
| Topic | Key facet | Sub-facet | |
|---|---|---|---|
| Query | "what is creatine?" (benefits, dosage, safety) | "when should you take creatine?" | "can I take creatine at night?" |
| Format | Blog article | YouTube long-form | Short / Reel |
| Note | The pillar. Covers the whole topic on one page | One angle of the topic, explored properly on video | One specific question, answered in 40 seconds or less |
| Title / hook | Creatine: What It Does, Benefits, Supplements & Safety | You're Taking Creatine at the Wrong Time (Here's How to Fix It) | "99% of people don't know when to take creatine, and it's not the time you think" |
Embed the video on the site page targeting the same cluster. Page and video now reinforce each other for the same queries, assembling the multi-format presence the SERP is starting to reward.
This is the article-video pairing run on instinct years ago, formalised. The pairs built back then are the videos still topping the analytics today.
The goal state is multiple placements on one results page: your page, your video, your image, and the AI answer drawing on all of them.
Each new placement also generates new query data in the platform property, which feeds the next revision of the map.
Hold multiple placements as text, video and social results for the same topics. The query data compounds, so go back to Step 2 and repeat the cycle.
Read exports at cluster level. That 149 phrasings constitute one demand pool is invisible in a query-sorted list.
Triage into Optimize, Embed, Produce. Not every cluster needs new content — positions 4 to 20 need a title and thumbnail rework first.
Prioritise the high-impression, low-click segments. Two hundred thousand impressions producing 87 clicks is unopened inventory, not failure.
Formalise the article-video pairing. Turning instinct into process is the framework's practical core.
Wait 28 days. Judging on week-one data means reacting to noise.
Check how machines understand your brand alongside it. Adding formats without stable entity understanding diffuses the effect — see The AI Entity Footprint Audit.
Read it with the creator citation data. Earned media accounting for 82% of AI citations is covered in 82% of AI Citations Point to Earned Media.
In an environment of declining posting volume — see Social Media Is Becoming Less Social — the case for posting less but building content that persists in search now has data behind it.
YouTube Studio shows YouTube Search — what people typed inside the platform. Platform properties show Google Search — the queries where your videos appear on the results page. Different populations and intents, and the second was a black box until July.
18,233 impressions on day one and more than 200,000 across 11 days against 87 clicks — extrapolating to roughly 6.7 million impressions and 2,900 clicks a year.
The feed rewards packaging, breadth and consistency and pays out fast; the SERP rewards query match and sustained relevance and pays out for years. In the case study, seven of the top ten videos were 2021–2023 search-first uploads.
Into Optimize (positions 4–20 with real impressions, rework packaging), Embed (the asset already answers well, place it on the matching site page), and Produce (real demand with weak or absent position, brief new content).
Three inputs — platform property queries, standard keyword research and Google Trends — organised by topic cluster rather than by platform, with each cluster carrying its demand evidence, dominant intent and fitting formats across platforms.
Week-one data looks like noise; clusters only became visible after a month, so give it 28 days. YouTube data is richest, while Instagram and TikTok footprints are smaller and skew navigational.
To apply what you just read to your own site, start with a free audit of where things are now.
A strategist replies within 24 hours on business days.

YouTube is testing Ask YouTube AI answers at the top of standard search results for how-to and open-ended queries. Here’s who sees it and what it means for video SEO.

A translated first keyword can make a foreign market look empty. An 80-query test shows how international SEO keyword research fails and how to build a market-entry map.

Daily briefings, Analytics connections, hreflang sitemaps, localization and WordPress plugin swaps: five ways an SEO lead uses Claude to cut repetitive work.