Why did my YouTube views drop?

Almost every "the algorithm is suppressing me" post is a misread analytics screen. A view drop leaves a fingerprint, and the fingerprint tells you which of five things actually happened. Work down the funnel in order — impressions, then click-through, then retention, then conversion — and stop at the first stage that broke.

First: are you sure it dropped?

Compare the video against its own first-N-days, not against your best upload and not against a public benchmark. A video published on a Tuesday in July is not comparable to one that went out before a holiday weekend. And a video that is four days old has not "failed" — YouTube is still deciding. Low numbers on a brand-new upload are the default state, not a diagnosis.

The one number that means nothing on its own: subscriber count. The one that means almost everything: how this video performs relative to your other videos of the same format. Shorts and long-form live in different worlds; comparing them produces nonsense.

Stage 1 — Impressions fell

If YouTube is showing your video to fewer people, the click-through rate is a red herring. Check impressions first, in Advanced Mode, on the traffic-source breakdown.

• Browse and Suggested collapsed, Search held steady. The recommendation system cooled on the video. That usually follows weak early retention or a topic that didn't hold a broad audience — the fix belongs to the next video, not this one.
• Everything fell at once, across the channel. Look at seasonality before anything else. Audiences leave in summer, during holidays, during major events. Your channel is not broken; your audience is at the beach.
• An old video stopped getting impressions. Search demand moves. The phrasing people used a year ago isn't what they type now, and your metadata is still speaking the old language.

Stage 2 — Impressions held, clicks didn't

This is the packaging stage, and it's the one creators most often misdiagnose. If YouTube is showing the video and people aren't clicking, the title and thumbnail are the bottleneck — full stop.

Ignore the "good CTR is 4–10%" advice you'll find everywhere. It's folklore. The only comparison that means anything is your own median: this video's CTR against the CTR your other videos earn, in the same format, at similar impression volumes. A 3% CTR is excellent on a channel whose median is 2.2% and alarming on one whose median is 6%.

Note the trap: raising CTR with a promise the video doesn't keep is a net loss. Clicks spike, retention craters, and YouTube's system — which optimises satisfied watch time, not clicks — suppresses the video harder than before. The title and thumbnail must make one honest promise the first thirty seconds pay off.

Stage 3 — People clicked, then left

Open the retention curve and find the first cliff. Then — and this is the step almost nobody takes — go and watch what your video is actually doing at that timestamp.

• A steep drop in the first 30 seconds, with an intact curve afterwards. That is a packaging-promise mismatch, not a pacing problem. The people who clicked expected something the opening didn't deliver, so they left; the ones who stayed are enjoying it.
• A slow, steady bleed. Pacing. Cut the tangents, tighten the middle.
• A cliff in the middle. Something specific happened there — a sponsor read, a slow section, a topic switch. Watch it. The answer is usually obvious once you look.

The uncomfortable truth about retention: you can't fix a published video's content. What you can fix is the promise on the outside, so the right people click in the first place — and the hook on the next upload.

Stage 4 — They watched, and didn't subscribe

Real traffic, healthy retention, flat subscriber growth. This is a conversion problem, and it's usually a search-traffic channel: one-off visitors got their answer and left satisfied. They'll subscribe when the video makes it clear what they get next week if they do. That's a value proposition, not a louder "smash that subscribe."

The rule that saves you from most of this

Change one thing, wait one to three weeks, and compare against your other videos over the same window — not against what the video was doing before. Channels swing for a dozen reasons that have nothing to do with your edit. If you change a title on Monday, the channel gets a seasonal bump on Wednesday, and you credit the title, you have learned something false and you'll repeat it.

Also worth reading: how Wedio differs from vidIQ and TubeBuddy · What Wedio is