YouTube keyword research, the full guide for marketers

Most YouTube keyword research starts in the wrong app. Someone opens a paid tool, exports two hundred keywords sorted by volume, and calls it a content plan. The videos ship, the views don't come, and the tool gets blamed for a problem the tool didn't cause.

I run the research side of client channels at Market Correct, and after 12+ years across 400+ clients of deciding what earns budget, the finding I keep coming back to is that the order of operations matters more than the tool. This is the guide I'd hand a marketer on day one. Each section links to a longer post where I've gone deeper. Treat this as the map.

What YouTube keyword research is, and how it differs from Google

YouTube keyword research is finding out what people type into YouTube's search bar, how much pull each phrase has, who already ranks for it, and which treatment of the topic is actually getting watched. It happens before the shoot, so the video is built against a query with real demand and a gap you can fill. If you come from web SEO, three things are different.

1. Search is one discovery path. Often it's not the biggest one. Suggested videos and the home feed do a lot of the work, so a keyword tool only ever describes one slice of how a video gets found.
2. Every volume number is a proxy. YouTube doesn't publish search volume anywhere. Every figure you've seen came from a third-party tool estimating it, and the tools disagree because they estimate from different data. Google Keyword Planner measures Google demand, not YouTube demand, so it isn't a stand-in either.
3. Packaging is the product. A title tag and a meta description are set-and-forget. On YouTube the hook in the first 15 seconds, the title construction, and the thumbnail decide whether demand converts to a watch. The keyword tells you what to make. It can't tell you how to frame it.

The workflow in six steps

This is the condensed cut of the sequence I run before any client channel gets a content plan. The long version, with what to look for at each step and a tools-per-step table, is my step-by-step guide to YouTube keyword research. Run it in this order.

1. Pull real queries from YouTube autocomplete. Type your topic into YouTube's search bar, write down every completion, then add a letter after the topic and harvest the next set. Twenty minutes usually produces 30 to 50 candidates, closer to YouTube intent than anything a web tool seeds you with.
2. Put volume numbers on the shortlist. Run the candidates through a tool that reports YouTube search volume so you're ordering by interest instead of by gut. VidIQ's native score, OutlierKit's keyword data, Keywords Everywhere inline, or Ahrefs in YouTube mode if you already pay for it. Volume ranks ideas against each other. It doesn't forecast views.
3. Read the competition before you commit. For each surviving query, look at who currently ranks. Channel size, video age, and production quality tell you whether there's a gap. Two-year-old videos from mid-sized channels are an invitation. Fresh uploads from channels with millions of subscribers mean not first.
4. Find the outlier videos in the niche. An outlier is a video doing 3x to 10x its own channel's average. A big channel getting big views proves the channel is big. A mid-sized channel getting 8x its normal views proves the video did something right. This is the step most keyword research skips, and it decides what gets filmed on my client work.
5. Steal the framing, not the topic. Study the outliers' packaging before you script anything. The hook, the title construction, the thumbnail concept. You're learning why that angle beat the baseline, then applying the lesson to your own take.
6. Turn it into a plan a team can run. One table, one row per video, columns for target query, volume estimate, competition read, outlier reference, and planned angle. When a video flops you can see which column lied.

Most teams do step 2 with an expensive tool and skip steps 3 through 5. That's why most videos land in the 95% that go nowhere.

The tool stack by budget

Most tool roundups grade every tool on the same checklist and crown whoever pays the biggest commission. I sort them by the job you're hiring for, because finding title ideas is a different job from estimating volume, which is a different job from spotting outliers. The full grading of six tools by job, including Keywords Everywhere and the honorable mentions, is my best YouTube keyword research tools roundup. The stack comes in three tiers.

Free. YouTube autocomplete, Google Trends, the Ahrefs free tool

You can get surprisingly far on no budget. YouTube's own search bar autocompletes the real queries people type. Google Trends has a YouTube Search filter that shows relative interest over time, so you can read seasonality and compare two topics head to head. Ahrefs publishes a free YouTube keyword tool that returns keyword ideas and volume estimates without a paid seat, though the volume is a web-derived proxy. What you give up is volume precision, competition scoring, and any outlier data. Run this first and upgrade once the guessing starts costing you shoots.

Paid web tool you already own. Ahrefs in YouTube mode

If you run SEO, you already live in Ahrefs, and it mostly works for YouTube. Keywords Explorer has a search-engine selector you can switch to YouTube, the filters for matching terms, questions, and word count carry over, and it wins on exportability (CSV, API). Where it stops is everything YouTube-specific. The volume is a web clickstream proxy, the Keyword Difficulty score is calibrated for Google's link-based ranking, and there's no outlier detection, no hook or thumbnail context, and no RPM data. For occasional title and tag ideas, it's enough. For working out why a competitor's video took off, it isn't. I wrote up how to use Ahrefs for YouTube and where it falls short, including a fast four-step Ahrefs-to-YouTube workflow.

YouTube-native. OutlierKit, VidIQ, TubeBuddy

OutlierKit is a research SaaS built around one uncomfortable fact, that 95% of YouTube videos get fewer than 1,000 views, so the smart move is to decide what to film before you film it. Documented features are outlier video detection, a low-competition keyword finder, high-RPM keyword targeting, a niche finder with RPM ranges and competition level, hook strength and AI script analysis, trend prediction, Deep Research, and competitor channel analysis. It's the tool I use for step 4, overkill for cheap one-off tag suggestions, and the trial is 10 credits with no card. I graded it against a six-point framework any marketing team can reuse in my OutlierKit review for B2B marketing teams, including the credit math per tier and the gaps (API listed as forthcoming, no documented export, AI script scores are directional).

VidIQ is the broad creator suite. Keyword research runs on a YouTube-native search-volume score, and around it sit AI Coach prompts, thumbnail tools, channel audits, and a daily-ideas feed. There's a free plan, and it's the easiest tool to live in day to day. It's lighter on pure outlier research, which it touches through trends. TubeBuddy is a YouTube Studio extension first. Tag suggestions, bulk processing across a back catalog, and A/B thumbnail testing are its strengths. Its Keyword Explorer is lighter and its volume read is the weakest of the major tools. They're complements wearing competitor costumes, and the row-by-row breakdown is my VidIQ vs TubeBuddy comparison.

Read the row that matches your bottleneck, not the column with the most yeses.

How to read YouTube search volume

Every YouTube volume figure belongs to one of two families, and the family matters more than the tool's brand. YouTube-native estimates are modeled from signals on the platform. VidIQ's score works this way, and so does OutlierKit's keyword data. Web-derived proxies borrow volume from Google-style clickstream data and map it onto YouTube queries. Ahrefs works this way. Native numbers are the better read for choosing between topics. Proxies are fine for a first pass, especially if you already pay for one.

Use any of these numbers comparatively, never predictively. Rank a shortlist, spot which phrasing of the same idea has more pull, confirm a niche has any demand at all, read seasonality. Don't forecast views with it, don't compare figures across two tools (different estimation methods make that meaningless), and don't greenlight a video on volume alone. The full explainer, including three free ways to check volume in about ten minutes, is my guide to YouTube search volume.

How to pick a title-shaped keyword

A keyword only counts as a target once you can build a title around it. Until then it's a topic. Turn broad seeds into title-shaped queries before you look at a volume number.

1. Start from autocomplete phrasing, not your own. The completions are already worded the way people search, usually a question or a "how to" plus a qualifier.
2. In Ahrefs, filter by questions, matching terms, and word count. Those filters carry over cleanly from web SEO and turn a raw dump into queries a title can actually contain.
3. Group variants under one video. Ahrefs's Parent Topic view maps a dozen keyword variants to the single video that should target them. Do it by hand in a spreadsheet on the free stack.
4. Check which phrasing has more pull, then let the outlier videos in that niche show you the title construction that beat the channel baseline. The keyword decides the topic. The outlier decides the framing.

Common mistakes

• Starting in the paid tool. The seeds come from YouTube's search bar. Anything a web tool seeds you with is one step removed from YouTube intent.
• Reading volume as a forecast. It's the number every tool puts in the biggest font, and it says the least about whether your video gets watched.
• Chasing the high-volume keyword ten massive channels already own. A mid-volume keyword nobody has served well is the better bet. This is the single most common mistake I see in creator research.
• Using Google Keyword Planner as a stand-in. It measures Google demand. People type differently into the two search bars.
• Building the plan on TubeBuddy's volume numbers. It's the lightest read of the major tools. Use TubeBuddy for tags and bulk cleanup.

Who should use what

Here's the routing I'd give a client, the same evidence-first read I bring to B2B performance marketing on the paid side.

Prices move often enough that I'm not printing them here. The OutlierKit review and the VidIQ vs TubeBuddy comparison carry the tier breakdowns, and the vendor pricing pages are the source of truth before you buy.

The bottom line

YouTube keyword research is a sequence. Seeds from autocomplete, volume to rank them, competition to filter them, outliers to prove the angle, packaging lessons from the winners, and a one-table plan at the end. Every volume figure is an estimate, so let it rank your ideas and never let it promise you views. The video that wins usually wins on the angle, and volume data has nothing to say about angles.

You don't need the expensive tool to start. Run the sequence on the free stack, keep Ahrefs if you already pay for it, and add a YouTube-native tool the moment a step is slowing you down. That's how I'd spend a client's budget, so it's how I'd spend yours.