IntermediateLesson 11 of 206 min read

Finding an Audience: Search, Suggested Videos and Browse

Score questions people ask, understand the three main doors to a video, and read a traffic mix without chasing the crowd.

In this lesson you will

  • Name the three main doors to a video
  • Score search phrases by hand
  • Read a traffic mix
  • Choose a niche question over a crowded topic

About 30 minutes

Who is going to find this?

Kofi has published eight videos, and one of them, about pinching basil, gets a slow, steady trickle of views every week without him doing anything. Another gets a burst on the day he uploads it and then goes quiet. Abena asks, “Why does one keep going and the other stop?” The answer lies in how people find videos. Understanding it at a simple level changes what you make next. Everything here is made up for teaching, and the lesson hedges on features, because the site changes how it shows videos from time to time.

Three main doors

Door What it means
Search Someone types a question and picks your video
Suggested Your video is offered next to or after another
Browse It appears on a home page or subscription feed

YouTube Help describes these as traffic source types in the Reach part of Analytics: viewers can find your videos from inside YouTube, such as search, suggestions and browse features like the home page and subscriptions, or from external sources, such as a link on a website. Each door needs a different kind of video. A search viewer has a question now, so the video must answer it plainly. A suggested viewer has just watched something related, so your video must feel like a natural next step. A browse viewer was not looking for anything, so the picture and title must make them curious, and your regular viewers must like what you make.

Looking for questions people ask

For a small channel, search is the easiest door to work with, because you can see questions in advance. Type a topic into the search box and note the suggestions that appear as you type, the questions people ask in comments on similar videos, and the questions your friends ask you. This is simple keyword research. There are paid tools that estimate search counts. This course makes up its numbers, and the program below shows the thinking with six invented phrases.

phrases = [
    ("why is my basil dying", 2400, 180, 5),
    ("how to pinch basil", 900, 40, 5),
    ("best herbs for beginners", 5200, 1500, 4),
    ("balcony tomato pots", 1800, 220, 5),
    ("herb garden", 22000, 9000, 3),
    ("basil leaves drooping", 3100, 260, 4),
]
rows = []
for phrase, searches, videos, fit in phrases:
    score = searches * fit / (videos + 1)
    rows.append((score, phrase, searches, videos))
rows.sort(reverse=True)
for score, phrase, searches, videos in rows:
    print(phrase + ": score " + str(round(score)) + " (" + str(searches) + " searches, " + str(videos) + " videos)")
print("biggest topic:", max(phrases, key=lambda p: p[1])[0])
print("best score:", rows[0][1])
views = {
    "Save your basil": (620, 180, 240, 40),
    "Three herbs to start": (410, 520, 300, 60),
    "Pinching tips": (150, 40, 55, 5),
    "Pot sizes explained": (300, 280, 190, 30),
}
print("order: search, suggested, browse, external")
total = [0, 0, 0, 0]
for name, v in views.items():
    s = sum(v)
    shares = [round(x / s * 100) for x in v]
    print(name[:18] + ": " + str(shares) + " of " + str(s))
    total = [a + b for a, b in zip(total, v)]
grand = sum(total)
print("channel search share:", round(total[0] / grand * 100), "percent")
print("channel suggested share:", round(total[1] / grand * 100), "percent")
how to pinch basil: score 110 (900 searches, 40 videos)
why is my basil dying: score 66 (2400 searches, 180 videos)
basil leaves drooping: score 48 (3100 searches, 260 videos)
balcony tomato pots: score 41 (1800 searches, 220 videos)
best herbs for beginners: score 14 (5200 searches, 1500 videos)
herb garden: score 7 (22000 searches, 9000 videos)
biggest topic: herb garden
best score: how to pinch basil
order: search, suggested, browse, external
Save your basil: [57, 17, 22, 4] of 1080
Three herbs to sta: [32, 40, 23, 5] of 1290
Pinching tips: [60, 16, 22, 2] of 250
Pot sizes explaine: [38, 35, 24, 4] of 800
channel search share: 43 percent
channel suggested share: 30 percent

Each phrase has a made-up count of monthly searches, a count of videos already about it, and a fit mark from one to five for how well Kofi can answer it. The program scores each phrase as searches times fit, divided by the number of videos plus one. The biggest topic by searches is herb garden, with 22,000 searches. But it ranks last. It has 9,000 videos already, and a small channel would be lost among them. The best score belongs to how to pinch basil: only 900 searches, but just 40 videos and a perfect fit. This is the lesson of a niche. A small, specific question that few have answered well often beats a huge, crowded one.

The formula is crude, and you should say so. The number of videos is a stand-in for competition, but it ignores quality: ten excellent videos are harder to beat than a hundred poor ones. The search counts are estimates. Treat the score as a way to compare ideas on one page and not as a forecast.

Reading your own traffic mix

The second part of the program looks at where four of Kofi’s videos got their views. Each line shows the shares from search, suggested, browse and external, in that order. Pinching tips is led by search, at 60 percent, which fits a “how to” question. Three herbs to start leans on suggested videos: 40 percent. Across the channel, search brings 43 percent of views and suggested videos 30 percent.

There is no ideal mix to aim for. A channel of practical lessons usually leans on search. A channel of entertainment leans on suggestions and browse. What matters is that you notice which door is open and make more of what works there. If a video gets almost all its views on release day from your subscribers and then stops, it did not find a door beyond your own audience.

What to do with this

  1. List ten questions that your target viewer might type. Use the search suggestions and real comments.
  2. Score them by searches, competition and fit, in the plain way shown above.
  3. Answer the best one first, with the question in the title and the answer in the first thirty seconds.
  4. Look at the traffic sources a few weeks later, and note which door each video used.
  5. Make a follow-up to the video that found a door, so that the viewer has somewhere to go next.
Common mistake: picking the crowd

Chasing the biggest topic because it has the most searches. The numbers are large, and so is the competition. Look for the question you can answer better than what is there.

Made-up numbers

The phrases, search counts and traffic numbers are made up. The scoring formula is a teaching device. Names of reports in YouTube Studio change, so check YouTube Help for the current wording.

Type a topic you know into the search box and write down six suggestions or questions. Guess which has the fewest good answers, then look.

Practice for this lesson

Check what you learned

Five quick questions. Pick an answer and you will see straight away whether it is right and why.

What are the three main doors for finding a video, as used here?

  1. Camera, light and sound
  2. Search, suggested videos and browse features
  3. Titles, tags and files
  4. Likes, shares and saves

Each door suits a different kind of video.

Why did the biggest topic rank last in the score?

  1. It has too few searches
  2. It does not fit anyone
  3. It is too short
  4. It already has thousands of videos, so a small channel would be lost

Crowded topics are hard to break into.

Why is the number of existing videos only a crude measure of competition?

  1. It ignores how good those videos are
  2. It counts only long videos
  3. It is always exact
  4. It is the same for every topic

Ten excellent videos can be harder to beat than a hundred poor ones.

What does a video led by the search door suggest?

  1. It is only for subscribers
  2. It has no audience
  3. It answers a question people type, such as a how-to
  4. It was removed

The pinching tips video was led by search.

What should you do after a video finds a door?

  1. Delete the other videos
  2. Change the channel name
  3. Stop uploading
  4. Make a follow-up so viewers have somewhere to go next

Build on what works.

Summary

  • Videos are found through search, suggestions and browse features.
  • Niche questions beat crowded topics for small channels.
  • Scores are crude: treat them as comparisons, not forecasts.
  • Look at which door a video used, and build on it.

Next lesson

Lesson 12 tests titles and thumbnails fairly, and explains why clicks alone can mislead.

Questions

Do I need paid keyword tools?

No. The search box suggestions, the questions in comments and your own audience give a good start. Paid tools can help later, but their numbers are estimates, so use them as one input among several.

Is search or browse better?

Neither is better in general. Search brings viewers with a clear question and tends to last. Browse and suggestions can bring bursts when a video suits a wide audience. Many channels use all three.

How long should I wait before judging a video's traffic?

Give it at least a few weeks, because search traffic builds slowly. Judging after a day mainly shows how your existing viewers reacted.

Progress is saved in this browser.