You already know how to shoot a good video. This is about what happens after export — the titles, descriptions, and tags that decide whether anyone finds it. By the end of this, you'll have a working system for writing metadata that actually matches how each platform's search and recommendation engine reads your content, instead of copy-pasting the same caption everywhere and hoping.
Why One Metadata Template Doesn't Work Everywhere
YouTube is a search engine first. TikTok is a recommendation engine that also happens to be searchable. Instagram sits somewhere in between, leaning harder on captions and audio than either. LinkedIn barely indexes video text at all — it's mostly about dwell time and engagement in-feed.
That means the same 60-second video needs different metadata thinking depending on where it lands. Not different videos necessarily — reformatting for different aspect ratios is its own workflow — but different words wrapped around it.
The good news: the underlying logic is consistent even when the mechanics differ. Every platform is trying to answer the same question — "what is this video about, and who should see it?" Your job is to answer that question clearly, in the specific format each platform expects.
The Three Things Every Platform Checks
Before getting into platform-specific rules, here's what's universal:
1. The first few words matter more than the rest. Whether it's a YouTube title truncated in search results or a TikTok caption cut off after two lines, front-load the actual subject of the video. Don't open with your brand name or a vague hook if you can help it.
2. Keywords need to appear where humans actually read them. Stuffing tags with terms nobody sees does very little on modern platforms. Titles, opening lines of descriptions, and on-screen text carry more weight than hidden metadata fields.
3. Consistency between what you say and what's on screen matters. If your title promises "how to fix a leaky faucet" and the video is about faucet shopping, platforms notice the mismatch through watch time and skip rate, even if they can't literally "read" the video.
YouTube Video SEO: Titles, Descriptions, Tags
YouTube is still the closest thing to traditional SEO in the video world, which is why most creators default to YouTube habits everywhere else — a mistake we'll get to.
Titles: Put your main keyword in the first 40-50 characters. "How to Edit Vertical Video for TikTok and Reels" beats "My Editing Process (Full Tutorial!)" every time, because it tells both viewers and YouTube's search index exactly what's inside.
Descriptions: The first two lines show up in search results and suggested videos, so treat them like a mini pitch, not an afterthought. Repeat your primary keyword naturally once or twice, then use the rest of the description to add context, chapters, and links — this is also where you can mention related resources, like your captions and subtitles workflow if the video touches accessibility.
Tags: Tags matter less than they used to, but they still help YouTube understand adjacent topics. Use a mix of broad ("video editing") and specific ("vertical video editing for TikTok") tags. Skip the temptation to tag with unrelated trending terms — it doesn't fool the algorithm and it can hurt trust signals.
TikTok SEO: Captions Do the Heavy Lifting
TikTok doesn't really have "descriptions" and "tags" the way YouTube does — it has one caption field, plus on-screen text and spoken audio that the app can now parse.
That changes the strategy:
- Write the caption like a mini title, not an afterthought. "3 mistakes killing your reach" performs better for discovery than emoji strings or vague hooks.
- Say your keyword out loud in the video. TikTok's algorithm reads spoken words, so if the topic is "budget travel tips," say that phrase somewhere in the first few seconds instead of only implying it.
- Use on-screen text as a second metadata layer. Text overlays are indexed too, so your hook text can double as a keyword placement.
- Hashtags are for categorization, not magic keywords. Two to four relevant hashtags is plenty. This site already covers hashtag strategy in depth, so we won't rehash it here — the point for SEO purposes is that hashtags support discovery, they don't drive it the way captions and spoken keywords do.
Instagram and LinkedIn: The Quiet Middle Ground
Instagram Reels blend YouTube and TikTok logic. Captions matter, alt text matters more than people realize, and the first line of your caption gets the same "truncated preview" treatment as a YouTube description. Write it like a headline.
LinkedIn is different again — it rewards native uploads and watch time far more than keyword-optimized text. A strong first sentence that stops the scroll matters more than a keyword-stuffed caption. Treat LinkedIn metadata as a formality, not a growth lever.
A Quick Comparison: One Video, Four Platforms
Say you made a 90-second video: "How to batch-schedule a month of content in one afternoon."
- YouTube title: "Batch Scheduling Content: Plan a Month in One Afternoon" — keyword-forward, clear promise.
- TikTok caption: "I plan a month of posts in one sitting — here's exactly how" — conversational, but the keyword phrase "plan a month of posts" still appears early and gets said aloud in the video.
- Instagram caption: Similar to TikTok, but with a stronger first line since Instagram truncates faster: "Batch scheduling saved me 10 hours a month."
- LinkedIn caption: A short, direct hook aimed at professionals: "Most content calendars fail because of this one step." No keyword stuffing — it's built for scroll-stopping, not search.
Same core message, four different metadata approaches. This is the part of cross-platform publishing that gets skipped when creators focus only on formatting and posting speed — and it's a big reason two accounts can post similar content with very different discoverability. If you're wondering whether posting the same video everywhere actually hurts you, the content itself usually isn't the problem — thin or copy-pasted metadata is.
Building This Into Your Workflow
Writing five versions of metadata for every video sounds like a lot of extra work, and if you're doing it manually for dozens of clips a month, it is. A few ways to make it sustainable:
- Write metadata as part of your atomization process, not after the fact. When you're already breaking one video into a week of posts using something like the content atomization framework, draft the platform-specific captions and titles at the same time you're cutting the clips.
- Keep a swipe file of keyword phrases for your niche, organized by platform tone — formal for YouTube/LinkedIn, conversational for TikTok/Reels.
- Track what's actually working. Metadata experiments only mean something if you're measuring outcomes. Comparing view-through and search traffic across channels is exactly what cross-platform video analytics is for.
- If you're managing this across a team, build metadata review into your existing approval workflow so titles and descriptions don't get rushed at the last minute before publishing.
- Tooling helps here more than people expect. A good distribution tool can store platform-specific title and caption templates so you're not rewriting from scratch every time — and switching tools later to get that feature often costs more in migration time than it saves.
The Real Takeaway
Video SEO across platforms isn't about learning a new trick for each app — it's about understanding that titles, descriptions, and tags are doing different jobs depending on where they live. Write for how each platform actually reads your content, not for how you wish it worked, and the same video will find its audience in more places without you reshooting a single frame.
