✦A small channel took the top spots
We recorded episode eleven with the screen up. Searching YouTube in incognito for gym interior design in Korean, the top of the page was filled with one company's content. The channel belongs to Inhwan Kang.
The odd part was that the view counts were not high. The top of the results was more or less taken anyway. This note is about how that was possible, and why a small view count can be fine.
For context, this channel was built together from the start. Nothing was reworked from existing videos with new titles or thumbnails. It began with opening the channel, and it started appearing at the top roughly two months later.
✦Is there SEO on YouTube
I asked, and the answer was that when people have a purpose they search on YouTube too, and making your brand's video appear at that moment is the same work, so yes. Then he added something.
When someone searches on YouTube with a keyword for a product they want to buy or something they are curious about, and our video appears, I genuinely feel the conversion rate to revenue is very high.
YouTube videos also appear in Google results. Search my podcast name on Google and the YouTube videos come up. So YouTube is less a separate platform than another entrance to search. That connects to what I wrote in the note on AI search.
✦How a small channel flipped the results
Other brands already held the top when he started. There was a specific basis for believing he could win anyway.
They were very old videos, and the channels were small. Being old means that brand is not publishing content now.
That scene shows up constantly in search too. A position at the top where updates have stopped is close to an empty position. So he attached three things. Freshness, putting the target keyword in the title, and publishing several pieces of that content.
There was also a read on demand. A gym fit out costs tens of millions of won at minimum and sometimes reaches into the hundreds of millions. Which means people research a great deal before deciding. And in video, a person shows their face and explains, which attaches trust.
Worth adding, he had run a gym before but had not properly started the interior business yet. Starting on YouTube was effectively starting the business.
✦Title, thumbnail, description, tags
The most practical stretch. The conclusion first: the title comes first. YouTube's own help center says the same.
Your video's title, thumbnail, and description are more important metadata for searchability.
What about tags? The same document says tags are useful when a word is commonly misspelled, and play a limited role otherwise. It also warns that putting excessive tags in the description is a policy violation.
Put the target query in the title
If you want to appear for a specific search, the title is the first priority. Leave it out and the rest will not fill the gap.
Judge the thumbnail on click through rate
The thumbnail's job is getting clicked once you have appeared. Whether to put a keyword on it depends on whether that raises the click through rate.
What you say is data too
YouTube reads the script through automatic captions. Whether the keyword you want to rank for is actually spoken in the video counts as well.
Do not agonize over tags
Time spent deciding how to fill tags is better spent on the title and description. YouTube has said outright that their role is limited.
The last point matters. Ten videos with a tidy title and description, even if the rest of the setup is imperfect, produce more data for YouTube to read than one video you poured everything into. Volume decides it in the end.
Planning website and YouTube together?
A video ranking and a site ranking are different jobs, and designed together they lift each other. Send me your site and target market, and I'll tell you whether I can help.
Send me a note✦Viral or search, pick one
The strategies fork. Content aimed at spreading to many people and content aimed at being found through a specific query are designed differently.
Something like gym interior design is very niche. In that case I think you give up the viral element somewhat.
Instead you design for what the person searching that query wants to know, and for substance, meaning content that creates trust. View counts come out low. Conversion is good.
I checked on my own channel. Opening traffic sources in the YouTube Studio analytics tab shows the queries bringing search traffic. Mine were arriving on terms like SEO optimization, GEO, and GEO marketing. Which means the direction is right.
He gave the counterexample too. A channel that ran YouTube ads early on does gain subscribers, but unrelated people arrive, and since they come in and leave, channel value can actually drop. Numbers manufactured because you needed a number to show become noise.
Press this and my new posts show up more often in your Google results.
✦Why a hundred views means something
I asked honestly. There are stretches where the view count feels small. His answer:
Those hundred views are only people who are genuinely interested, so from the moment you feel the interaction, that is the drive to keep making them.
He brought up his own experience. The goal from the start was for clients to come from watching the channel, and as soon as he passed a hundred subscribers a division head at a startup got in touch and he went to a meeting. The terms did not work out, but the other side arrived already persuaded by the content.
A meeting where the other side asked for it, and a meeting where I raised that client's temperature myself, are worlds apart.
The comparison by numbers stayed with me. In real estate there are brands with two to three hundred thousand subscribers that struggle to fill thirty seats at a weekly offline seminar, while a client with five to six thousand subscribers has more than thirty people standing through seminars twice a week.
That client went as far as saying thirty thousand subscribers would be enough and he does not want more, because beyond that comes an audience he cannot manage and noise with it. So he does not cover beginner material at all. You decide through the content what kind of person you want to meet.
✦How to keep going with a day job
This part is about sustaining it. I plan, film, and edit alone alongside client work, so it slips easily. The answer was clear.
Put it on a timetable
Like a school schedule, fix planning, filming, and editing to specific days and times. If first period is English you sit through English even when you are tired.
Film in batches
He planned three or four ten minute pieces across three hours on Monday afternoon, then filmed all four on Tuesday morning. Reading the script takes ten to twenty minutes per piece, so two or three hours covered it.
Stop when the time is up
Editing stops after two hours even if the captions are not done. Planning and filming stretch forever if you let them. If a video that took thirty hours gets a hundred views, you give up before you make ten.
Buy equipment later
A smartphone is fine. If you have around three hundred thousand won to spend, spend it on a microphone rather than a camera. Sound quality drives immersion more than picture quality.
He gave a benchmark for time as well. Publishing a roughly ten minute long form video once a week alongside shorts cut from it, he had not seen a case in his experience where there was no interaction with a customer within six months. The path to an enquiry does have to exist, meaning a contact or a website link in a pinned comment or the description.
And the closing advice. Worry about it after you have published a hundred videos. That is roughly how much data it takes before you can judge success or failure.
✦Niche is the answer
Something late in the conversation matched my own work. A cosmetic surgery clinic saying it does eyes, nose, breasts, and liposuction inspires less confidence than one saying we only do eyes, and gathers a better targeted audience. YouTube works that way and so does search.
What I do on Google results is also territory capture. Taking the narrow ground somebody has vacated, and taking it properly, is faster. That is written out with the numbers in the note on long tail keywords.
One more thing. He talked separately about why he does not exaggerate in content. Someone who arrives through search has already looked into the field, so overstatement is visible immediately.
How I work is on the services page, and the other episodes are in the episode list.
Sources and notes
- YouTube Help, Add tags to YouTube videos. Read it
- The Insung Song SEO Podcast, EP 11, with Inhwan Kang, published 6 August 2026. Recorded in Korean. Watch it
The quotations are Inhwan Kang's words from the episode, translated from Korean by me, together with sentences from YouTube's official documentation. Search results and channel data were checked on screen together at the time of recording in August 2026 and may have changed. YouTube's ranking logic is not published, and the judgments here rest on the guest's working experience and the official documentation.
✦Common questions
The title. YouTube's own help center states that title, thumbnail, and description are the more important metadata for search. If the query you are targeting is not in the title, the rest will not make up for it. Judge the thumbnail on click through rate rather than keywords.
Their role is limited. YouTube states that tags are useful when a word is commonly misspelled and play a limited role in search otherwise. Excessive tags in the description are a policy violation. Time spent agonizing over tags is better spent on the title and description.
The case in episode eleven is exactly that. Starting from a brand new channel, it began appearing at the top of the Korean query for gym interior design about two months later. The reasoning was that the videos already at the top were old and their channels had stopped publishing, so freshness, keyword in title, and publishing volume were enough.
If those hundred arrived by searching, it means something different. The guest received a meeting request at around a hundred subscribers, and described a client with five to six thousand subscribers filling offline seminars better than channels with two to three hundred thousand. The link to contact you does need to be in the description or a pinned comment.
I help companies grow in search, mostly on Google, in Korea and across borders. Since 2021, the field has taught me to look at content and structure together rather than chasing one technical fix.
