Shorts vs. Long-Form on YouTube: What the Data Actually Shows

Every creator eventually asks the same question: should I be putting effort into Shorts, or is that time better spent on long-form? The honest answer is "it depends on the channel," which is unsatisfying but true — and the only way to get past that non-answer is to actually pull the numbers for channels in your niche instead of relying on general advice from a thread on Reddit.

Setting up a fair comparison

The mistake most people make when comparing formats is comparing raw view counts. Shorts naturally get more views because they're pushed harder by the algorithm and take fifteen seconds to consume instead of fifteen minutes. A fairer comparison looks at a few normalized metrics:

• Views per subscriber — controls for channel size


• Engagement rate (likes + comments / views) — controls for the inflated view counts Shorts tend to get


• Watch-time proxy — harder to get directly, but average view duration if available, or comment depth as a rough substitute

Pulling channel-level data

To do this properly you need a decent sample — a handful of channels, their last 30-50 uploads split by format, with view and engagement data for each. Doing this by hand in YouTube Studio only works if you own the channel; for competitive or industry-wide analysis you need to pull public video metadata programmatically.

import requests



def get_channel_videos(channel_id, token):

    url = "https://ensembledata.com/apis/youtube/channel/videos"

    params = {"channel_id": channel_id, "token": token}

    res = requests.get(url, params=params)

    return res.json()["data"]



def get_channel_shorts(channel_id, token):

    url = "https://ensembledata.com/apis/youtube/channel/shorts"

    params = {"channel_id": channel_id, "token": token}

    res = requests.get(url, params=params)

    return res.json()["data"]

EnsembleData splits regular uploads and Shorts into separate endpoints, which is convenient here since you don't have to guess at video duration to classify format yourself. Full parameter references and response fields are in their API docs if you want to see what else comes back in the payload (upload dates, view counts, etc.).

What tends to show up in the data

Without pretending there's a universal answer, a few patterns come up often enough to be worth checking on your own dataset: Shorts tend to have a wider variance in performance (bigger hits, but also more duds), and channels that use Shorts as a top-of-funnel to long-form content generally see better long-form retention than channels running the two formats as unrelated content streams. But "generally" is doing a lot of work in that sentence — the only way to know what's true for a specific channel or niche is to pull the actual numbers and look.

If you're making a format decision with real budget or time behind it, spend an afternoon building this comparison for five or six comparable channels before trusting a blanket recommendation from anywhere, including this post.