Algorithm

The Home Feed: How Browse Features Pick Your Video

The home feed runs primarily on watch history, not the video someone is watching. Here is how Browse features choose what to show and how to earn a place in it.

Open the YouTube app and the first thing you see is the home feed: a grid of videos chosen for you before you have searched for anything. For a huge share of channels, this is the single biggest source of views. It is also the hardest surface to influence directly, because you are not answering a question someone asked. You are being placed in front of someone who was just browsing.

In your Analytics, this whole surface shows up under "Browse features." YouTube's own definition lumps several places together: the home page, the subscriptions feed, Watch Later, and Explore or Trending, plus other browsing surfaces. So when you see Browse traffic, it is mostly home but not only home.

The main signal: watch history

Here is the distinction that explains everything about the home feed. According to YouTube, the home feed relies primarily on a viewer's watch history, while suggested videos rely primarily on the video they are currently watching. The home feed has no "current video" to anchor on, so it leans on the longer pattern of what you have watched and enjoyed over time.

That changes the problem completely. In suggested videos, you win by being the natural next watch after a specific popular video. In the home feed, you win by being a strong match for the kind of viewer who already watches videos like yours, based on their history. You are not riding one video's coattails. You are being matched to a person.

How the system actually decides

YouTube has said its recommendation system pulls from over 80 billion signals a day: watch history, search history, subscriptions, likes and dislikes, "not interested," "don't recommend channel," and post-watch satisfaction surveys. The system has two stated goals, to help each viewer find videos they want and to maximize long-term satisfaction.

Underneath, two pillars do the work. Personalization compares your watch and search history against viewers with similar tastes. Content performance asks how a video is doing with the people it reaches. The performance side has three buckets YouTube names directly:

  • Appeal: did people choose to watch this, or did they ignore it or hit "not interested"? Note that YouTube's term here is appeal, not CTR.
  • Engagement: once they clicked, did they stick around?
  • Satisfaction: likes and dislikes plus post-watch surveys asking whether the time was well spent.

How to earn a place in the home feed

Because the home feed matches videos to viewers based on history, the work is less about any single video and more about being a channel the algorithm understands. A few moves make you legible:

  1. Build a clear identity. If your channel consistently makes one kind of thing, the system learns exactly which viewers to test you on.
  2. Win on appeal first. A thumbnail and title that people choose to watch is the entry ticket, because the home feed is competing for an idle viewer's attention.
  3. Hold them with engagement and satisfaction. A video that gets clicked and then satisfies feeds the signals that earn wider home distribution.
  4. Give the system room to test. Early performance with a small audience decides whether YouTube widens the reach, so your opening minutes and packaging do a lot of work.
  5. Keep viewers in your videos with end screens and cards, which strengthens the satisfaction signals YouTube watches. See chapters, cards, and end screens.

Why this surface feels so unpredictable

The home feed is where creators feel the algorithm is random, and there is a real reason. YouTube has acknowledged that factors outside your control affect reach: overall interest in your topic, competition from other channels, and seasonality. A video can be well made and still get muted distribution because the topic is crowded or the timing is off.

This is also where the "my video flopped" panic usually starts. Before you blame yourself, it is worth understanding how the home feed naturally produces a wide range of outcomes. We dug into the real reasons in why good videos flop.

Reading your Browse traffic honestly

You will see plenty of advice claiming a "healthy" channel gets some specific percentage of views from Browse. Ignore the exact numbers. There is no official ideal traffic mix, and any "X% browse is healthy" benchmark is opinion, not a YouTube statement. The right mix depends entirely on your format: an evergreen tutorial channel leans on search, a topical commentary channel leans on Browse and suggested.

What is worth tracking is your own trend over time. If your Browse share is climbing as your channel finds its lane, that is the system learning who to show you to. If you want the full breakdown of every source, your traffic sources, decoded covers what each one means and how to read them together.

The honest summary

The home feed matches videos to viewers based on watch history, judged on appeal, engagement, and satisfaction. You earn a place in it by being a legible channel with a clear identity, packaging that earns the choice to watch, and videos that satisfy the people who click. Some of the outcome is outside your control, which is normal. Build the channel the algorithm can understand, and the home feed starts handing you the right viewers.

Frequently asked questions

What signal does the YouTube home feed rely on most?

Watch history. YouTube has stated that the home feed relies primarily on a viewer's watch history, whereas suggested videos rely primarily on the video they are currently watching. The home feed matches a video to the kind of viewer who already watches similar content.

What is the difference between Browse features and suggested videos?

Browse features cover the home page, subscriptions, Watch Later, Explore or Trending, and other browsing surfaces, driven mostly by watch history. Suggested videos are recommendations next to or after another video, driven mostly by the current video. They are separate traffic sources in your Analytics.

Is there an ideal percentage of views from Browse?

No. YouTube does not publish an ideal traffic mix, so any "X% browse is healthy" benchmark is opinion rather than an official figure. The right mix depends on your format, and the useful thing to watch is your own trend over time.

Why do some good videos get little home-feed distribution?

Factors outside your control affect reach. YouTube acknowledges that overall topic interest, competition from other channels, and seasonality all influence how far a video travels, so a well-made video in a crowded topic or off-season can still see muted distribution.

Keep reading

AlgorithmAnalytics

Why Good Videos Flop (and What the Data Says)

A great video can still flop, and most of the reasons are in YouTube's own docs. Here is what the data actually says before you blame yourself.

February 21, 2026 4 min read
YouTube SEOAudience & retention

Chapters, Cards, and End Screens: Small Levers That Add Up

Chapters, cards, and end screens are small on-video features with strict rules. Used well, they nudge viewers to the next watch instead of the exit.

March 1, 2026 5 min read
AnalyticsPackaging

What's a Good YouTube CTR? The One Number YouTube Will Confirm

Everyone quotes a "good" YouTube CTR. YouTube publishes exactly one figure, and it is a wide range. Here is the official number and how to read it without fooling yourself.

May 7, 2026 5 min read