YouTube SEO Strategy

I start with keyword research. I open YouTube, type a broad topic, and note the autocomplete suggestions — those are real queries people are typing. For a recent video on fixing a leaky tap, I saw "leaky faucet repair" had roughly 8,000 monthly searches inside YouTube, while "leaky tap repair" had just 2,000. I ran the same phrases through Google Trends and confirmed the volume difference. I picked "leaky faucet repair" because it also matched Google search queries, which meant the video could rank in both platforms. Next, I write the title. I put the primary keyword in the first 60 characters. For that video, the title became "Leaky Faucet Repair in 10 Minutes | Easy DIY". The keyword is front-loaded, the length is 54 characters, and it includes a benefit (10 minutes) and a format cue (DIY). I avoid stuffing — one keyword, placed naturally. Then the description. I write the first two sentences as a mini-abstract that includes the target keyword and a promise. For example: "Learn a quick leaky faucet repair technique that takes 10 minutes. This step-by-step guide covers tools, safety, and common mistakes." The keyword appears in the first 25 words, which helps YouTube match the video to the query. I also add timestamps for chapters: "0:00 Intro", "1:30 Tools needed", etc. Each timestamp is an opportunity to rank for the sub-topic phrase. The thumbnail is the last thing I finalise. I use a high-contrast image with a single focal point — a close-up of the dripping tap with a red arrow pointing at the handle. I test two versions using TubeBuddy's thumbnail test on my low-traffic channel to see which gets higher click-through rate. Once the video is live, I monitor audience retention in YouTube Studio. If the average view duration drops below 60%, I note the drop-off point and adjust future videos to front-load the value. Retention is the signal YouTube weighs most after click-through rate. I also add cards and an end screen linking to a related video in the same playlist. That playlist is another optimisation: I group three to five videos on the same topic so that YouTube's algorithm sees channel-level topical coverage. This full process takes about an hour per video, and it has doubled my organic impressions since I started. I have written more about broader [SEO services](/seo-services/) and how they apply to video, but this specific path is what I follow every time.

The most common mistake I see is keyword stuffing in titles and descriptions. I have seen titles like "Best leaky faucet repair DIY fix leaky faucet easy" — that hurts readability and probably triggers a penalty. YouTube's algorithm is smart enough to parse natural language; one clean placement of the keyword is enough. Do not repeat the keyword more than once in the title. Second watch-out: ignoring the thumbnail. I have a video on my channel — a clear how-to on changing a car tyre — that had a decent title and description but a dark, low-contrast thumbnail. It got 200 impressions in the first week. I swapped the thumbnail for one with a bright yellow background and a red circle around the tyre jack. Impressions jumped to 1,800 the next week. Click-through rate is a major ranking signal, and the thumbnail is the lever you can pull fastest. Third: publishing without captions. Even if you do not upload a transcript, enable auto-captions and correct the errors. YouTube indexes the text, and Google uses it to understand video content. One client of mine saw a 15% increase in search traffic after adding a human-reviewed transcript to each video. Fourth: keyword cannibalisation within your channel. If you have two videos targeting the same primary keyword, they compete. I use a spreadsheet to track the target keyword for each video and avoid overlap. If I do need to cover the same topic again, I link between the videos in cards and end screens so YouTube sees them as complementary. Finally, do not treat tags as a primary signal. I used to spend 10 minutes stuffing tags; now I add five to eight relevant tags and move on. The real work is in the title, description, thumbnail, and viewer retention. For niche channels, such as a [law firm SEO](/law-firm-seo/) channel, the same principles apply but the keyword pool is smaller — you need to be more precise with long-tail phrases. In [b2b SEO](/b2b-seo/) contexts, I have found that video descriptions benefit from adding a link to a related white paper or case study, which signals depth to both the viewer and YouTube's algorithm.

When I first started optimising for YouTube, I believed tags were the secret weapon. I would collect 20 to 30 keyword variations, stuff them into the tag field, and assume my video was set. After six months, I went back and compared videos with heavy tag usage against videos with minimal tags (fewer than 10). There was no correlation with ranking. Tags are a weak signal; the description, title, and viewer behaviour matter far more. I also assumed that a longer description automatically helps. I once wrote a 1,000-word description for a video thinking more text meant more indexing. In reality, YouTube truncates the description after about two lines in search results, so only the first 120-150 characters are visible. If your keyword does not appear in that visible snippet, you lose click-through rate from the search result page itself. Now I keep the first two lines tight and keyword-rich, then add optional context below. Another mistake: I ignored end screens and cards. I thought they were spammy and that viewers would leave if I promoted another video. Actually, a well-placed end screen or card increases session watch time, which is a positive signal for YouTube. I tested this: on a channel with 50 videos, I added end screens to all existing videos. Over the next month, average channel-level watch time increased by 11%. The same applies to playlists. I used to publish standalone videos without grouping them. Now I create a playlist for every series or subtopic, and I link the playlist in the description. One edge case: I work with a [healthcare SEO agency](/healthcare-seo/) client whose YouTube channel covers patient education. Because the subject is sensitive, we cannot use clickbait thumbnails; instead, we rely heavily on accurate transcripts and detailed timestamps. That niche forced me to realise that thumbnail strategy must adapt to audience trust expectations. Finally, I thought publishing frequency was the main growth lever. I pushed out three videos a week, each with mediocre retention. The algorithm ignored them. When I dropped to one video per week but spent the extra time on retention analysis and thumbnail A/B testing, impressions grew by 40% in two months. Frequency matters less than retention. If you are in a competitive space like [legal SEO](/legal-seo/), this lesson is critical: one well-retained video beats ten average ones.