Podcast: Download (Duration: 51:03 — 58.7MB)
YouTube SEO works because search traffic is passive and consistent, which builds a subscriber base steadily rather than in spikes. Sam Oh shifted the Ahrefs channel to targeting search terms when it capped out at 1,000 to 3,000 views per video, and the channel now sits past 150,000 subscribers.
In this episode I spoke with Sam Oh, Director of Product Education at Ahrefs and the face of their YouTube channel. He started in ecommerce in 2009, sold that business, ran an SEO agency, and now leads SEO education for one of the best keyword research tools available.
Below is the whole approach: how to find rankable keywords, how to judge whether you can beat a channel, why niche focus matters, how to build retention bumps, and how it applies to an ecommerce store.
Get My Free Mini Course On How To Start A Successful Ecommerce Store
If you are interested in starting an ecommerce business, I put together a comprehensive package of resources that will help you launch your own online store from complete scratch. Be sure to grab it before you leave!
Table of Contents
Key takeaways
- YouTube search traffic is passive and consistent, which produces steady subscriber growth rather than spikes.
- YouTube volumes differ from Google because platform intent is different.
- Start narrow and expand outward. Channel relevance determines whether you can rank at all.
- Judge competitors by engagement, not subscribers. Big channels publishing to 300 views have spread too thin.
- Build retention bumps with moments worth rewinding to, then find them in your analytics.
- Never send viewers off YouTube. Ending a session hurts you more than the click is worth.
- Engagement to views should run 1% to 10%. Outside that range, something is inflated.
- Do not repurpose a blog post as a script. Writing styles and delivery are different.
Why YouTube SEO beats chasing viral videos
Search traffic accumulates steadily, which is what compounds a subscriber base.
Before the shift, the Ahrefs channel had roughly 12,000 subscribers and could only reach its existing audience, capping at 1,000 to 3,000 views per video.
The turning point was accidental. A video called SEO For Beginners started ranking for the term SEO, which redirected the whole strategy toward ranking deliberately.
The difference shows in the analytics. YouTube search traffic climbs like a steady mountain, and browse features traffic from the homepage spikes and drops with momentum.
Sam’s framing is that YouTube is a momentum channel more than a search channel, and search is what generates the reliable underlying growth.
How to do keyword research for YouTube
Start with broad seed keywords in a keyword tool, then validate them against YouTube itself.
Sam begins with terms tied to the product, like SEO, link building, or keyword research, and checks volume.
Then he searches those on YouTube and looks at how many views the top videos actually get.
The next check separates search traffic from audience traffic. Third-party channel analytics show whether a channel’s views come from its subscriber base or from search.
The final step uses your own data as a benchmark. Take a keyword you already rank first for, note the views it produces, and compare relative interest in a new term against it to estimate the traffic.
Why YouTube search volumes differ from Google
People arrive on YouTube with different intent than they bring to Google.
The two reasons people search YouTube are entertainment and learning something that benefits from visuals.
How to tie a tie works well on YouTube because the visual is the value. A niche technical SEO concept does not.
Business topics illustrate the gap. Plenty of people search YouTube to learn SEO, and the volume does not approach entertainment categories.
The workaround for a dry product is reframing. A cardboard box company will struggle with videos about cardboard boxes, and DIY projects using cardboard boxes have a real audience.
Why a new channel has to start narrow
Channel relevance determines whether YouTube will rank you for a topic at all.
Sam calls it channel authority. Because Ahrefs has so many videos on SEO, keyword research, and link building, a new video on those topics ranks in the top few positions within seconds of publishing.
The video then moves up and down for a period, settling based on how well it holds attention.
Expansion works in adjacent steps. Ahrefs ranks first for digital marketing, online marketing, and internet marketing from the same video, and ranks for ecommerce SEO because it is still SEO.
A video on making coffee would never rank for them, because the channel has no relevance to coffee and subscribers would not click.
His example of the right progression is vegan recipes to vegetarian recipes rather than vegan recipes to something unrelated.
How to judge whether you can outrank a channel
Look at whether a competitor’s subscribers are actually engaged rather than at the subscriber count.
Third-party analytics show whether a channel’s subscriber and view counts are growing or declining.
The tell is a large channel with weak recent performance. Millions of subscribers with new videos getting 300 views means they have spread themselves too thin and are not a real threat.
Small channels similar to your size are the realistic targets.
Content focus matters too. A channel publishing across everything in a category will be difficult to outrank broadly unless your specific video is exceptional.
The engagement ratio that reveals bought views
Divide likes, dislikes, and comments by view count to sanity check a competitor.
Videos earning views organically typically land between 1% and 10%.
Below 1% suggests bought views, since the views arrived without humans attached.
Well above the range for a video that does not seem to warrant it can also indicate purchased engagement.
Middling anomalies have innocent explanations too. Video embeds do not let viewers like or comment without clicking through, and paid ad traffic behaves similarly, which makes those videos poor competitive benchmarks.
How to structure a video that holds attention
Hook the viewer, explain why to keep watching, then get straight into the content.
Sam does not enforce a fixed intro length and does apply strict editorial rules to the script.
The cutting rule is specific. If 1% of the audience cares about something and 99% do not, it gets removed.
Length follows the content. Ahrefs rarely publishes under five minutes and sets no maximum, since padding kills engagement metrics.
He uses a teleprompter and writes by speaking aloud, so the script sounds like natural speech rather than written prose.
How to create audience retention bumps
Give viewers something worth rewinding to, then find those moments in your analytics.
The absolute audience retention graph shows how many people watched each part of the video, trending downward overall.
Bumps in that line mean people rewound, which pushes that segment’s count back up.
Sam’s example is a keyword modifiers list in a keyword research tutorial. He tells viewers to take a screenshot and move on, and the graph shows a clear bump there.
A sharp early drop from 100% to around 30% tells the opposite story. The intro is too long or boring, or the title misrepresented what the video delivers.
Once you find what produces bumps, reuse the pattern in future videos and test whether it repeats.
What to use instead of a talking head
Nobody stays engaged watching the same shot for an entire video.
B-roll is the baseline and works for any channel, simply because attention needs visual change.
Custom animations often explain a concept better than words, which is why Ahrefs uses them for things like how Google crawls the web.
Screen recordings work well for anything demonstrable.
Their production workflow keeps this efficient. Sam records, notes what he wants in a document, and the editor works from color-coded markers in the timeline without needing SEO knowledge.
Why you should never send viewers off YouTube
Session watch time counts, and pushing people to your website ends the session.
Session watch time is the total time someone spends on YouTube before closing it, and your videos contribute to it.
It is not directly measurable, and you can approximate it through card clicks and retention on the videos those cards lead to.
Ahrefs never tries to lead people off the platform. Anyone genuinely interested searches for the brand and finds them.
Their description carries one link to the homepage as a signature, with external links reserved for things like a template spreadsheet.
The exception Sam allows is a specific product being demonstrated, where linking is natural and helpful rather than a broad set of links to every category page.
What to put in a video description
Include your primary keyword and a brief overview of what the video covers.
The reason is comprehension. Search engines still struggle to understand video, so text context helps both YouTube and Google categorize it.
Do not paste your blog post. Sam recommends writing it fresh or pulling from the script, since it takes very little time.
Ahrefs also includes links to their own related videos in every description, which supports session watch time.
Why not to turn every blog post into a video
Some topics belong in text and never work on video.
Sam’s example is keyword cannibalization, a narrow technical topic interesting to a small subset of readers.
YouTube skews earlier in the learning journey. People click things like how to rank number one in Google rather than deep technical specifics.
Scripts also cannot be lifted from posts. Writing style and delivery differ enough that reading a blog post aloud sounds unnatural.
How to apply YouTube SEO to an ecommerce store
Research competitors first, then build a search traffic base before attempting anything creative.
For a coffee roaster, that means putting coffee into a keyword tool, reviewing the keyword ideas report, and finding where volume exists.
Cross-reference those against YouTube to find where you can realistically rank.
Then expand outward to terms that do not contain your main keyword at all. French press and Aeropress both reach the same audience without the word coffee.
The sequence matters. Search traffic builds a tightly focused audience first, and only once that exists does out-of-the-box creative content have people to reach.
Sam Oh’s answers on general SEO questions
On guest posting after Google’s comments about devaluing those links, Sam has not seen a decline, and reads the guidance as targeting spammy content published at scale rather than genuine contributions.
On hiring an SEO agency, research what they have actually delivered rather than the sales pitch, which requires knowing enough to recognize whether the links they build are quality.
On the disavow tool, Ahrefs does not use it. Google is generally able to identify obvious negative SEO attacks, and Sam would only consider it under a manual action, and not as a first response.
On scraper sites republishing your posts with links, ignore them entirely.
On anchor text for guest posts, write naturally and link where relevant. Partial match anchors work well because they read naturally in context.
On internal links, add them to every post. Search your own domain for the target keyword to find relevant existing pages, and check which of your pages carry the most link authority for the strongest internal links.
Frequently asked questions
Does YouTube SEO still work?
Yes, and it produces different results than viral content. Search traffic is passive and consistent, which grows a subscriber base steadily. Sam Oh’s channel analytics show search climbing like a mountain while homepage traffic spikes and drops.
How do you do keyword research for YouTube?
Start with broad seed terms in a keyword tool, then search them on YouTube and check what the top videos actually get. Verify whether those views come from search or from the channel’s existing audience.
Why are YouTube search volumes different from Google?
Platform intent differs. People come to YouTube to be entertained or to learn something visual, so how to tie a tie performs well while narrow technical topics that read better as text do not.
Should a new YouTube channel stay narrow?
Yes. Channel relevance determines whether YouTube ranks you at all. Expand in adjacent steps, like vegan recipes to vegetarian recipes, rather than jumping to an unrelated topic your subscribers will not click.
How do you know if you can outrank a competitor’s video?
Check whether their subscribers are actually engaged. A channel with millions of subscribers whose new videos get 300 views has spread too thin. Channels similar to your size are the realistic targets.
What is a good engagement rate on YouTube?
Likes, dislikes, and comments divided by views typically runs 1% to 10% for organically earned views. Below 1% suggests bought views, and unusually high numbers on an unremarkable video suggest purchased engagement.
Should you link to your website in YouTube videos?
Sparingly. Sending viewers off YouTube ends the session, which counts against you. Ahrefs keeps one homepage link as a signature, and Sam allows linking a specific product being demonstrated.
Can you reuse a blog post as a video script?
No. Writing style and delivery differ enough that reading a post aloud sounds unnatural. Some topics also belong only in text, since YouTube audiences tend to be earlier in the learning journey.
