How to grow a YouTube channel from zero: a 90-day plan
How to grow a YouTube channel from zero: a clear promise, Shorts and long videos, the metrics to watch and the rules for monetizing. A 90-day plan to follow.
A new channel starts with one resource: the chance to learn fast what works. This guide promises no timelines or growth figures, because no reliable data guarantees them. It proposes a method for the first three months and the few numbers to watch so you can decide what to keep.
A note on sources. For the rules we use YouTube's official pages. For numbers on how Shorts perform we found only one study, from a tools vendor, Metricool, and we quote it with that label.
Where do you start when the channel is new?
You start with a promise in one sentence: who the channel is for, what problem it solves and why they should watch you. With a specific promise you pick three recurring topics and begin to publish, instead of chasing every subject that seems fashionable.
Write the sentence before you record: “I help [who] to [do what] with [which method].” It is the same logic we use for courses, explained in the guide to creating and selling an online course. If you can't finish it, the channel is still a topic, not a product: the first videos exist to find it.
How to choose your first three topics
Pick three topics you know well and for which you already have real questions, for example those friends, clients or colleagues ask you. For each, list five specific questions: that gives fifteen video ideas that start from someone's need, not from whatever “works” at the moment.
Put the topics to the test with your first ten videos, without waiting to be perfect. After ten releases you have enough data to see which topic holds viewers best and which doesn't interest them. If a topic doesn't work, it is not a failure: you found out in ten videos what you would otherwise find out in a year. And remember a channel serves a specific audience: a narrow topic with returning viewers beats a broad one with occasional visits.
What does YouTube say about how recommendations work?
It says which signals it considers, with no figures or weights: viewing behaviour, likes and dislikes, subscriptions, satisfaction surveys and channel reputation and quality, with input from external evaluators. Recommendations appear on the Home page and in Up Next.
That comes from YouTube's “How YouTube Works” page on recommendations, which carries no date. The page says nothing about posting frequency, subscriber count or click-through rate: that is why you cannot lean on “rules” someone claims to have discovered. What you can do is watch your own data and adjust.
Shorts or long videos: which is better?
Shorts are mostly for being discovered, long videos for explaining and building the relationship: that is an editorial judgement, only partly supported by available data. In Metricool's study of 799,718 videos from 71,177 accounts, the Shorts feed generates 61.54% of the sample's views and subscribers 11.31%.
See the data as a table
| Source | Share |
|---|---|
| Shorts feed | 61.54% |
| Subscribers | 11.31% |
| YouTube search | 8.8% |
| Related videos | 5.77% |
| Playlists | 5.05% |
| Channel page | 3.12% |
The same page warns that YouTube changed how Shorts views are counted on 31 March 2025, so comparisons across years span two definitions. And it gives no rate at which Shorts viewers become subscribers. Use Shorts to attract, but link every Short to a long video and a clear invitation, as we explain in content repurposing for creators.
| Shorts | Long videos | |
|---|---|---|
| Getting discovered by new people | Yes | Partly |
| Explaining a topic in depth | No | Yes |
| Building a relationship with followers | Partly | Yes |
The editors' judgement, not a measurement: Metricool's data does not say how many Shorts viewers become subscribers.
How many videos should you post a week?
There is no official number. In Metricool's sample, channels that posted 2 to 4 Shorts a week got the best results per Short, but it is a vendor's figure and the page itself says causation is not shown.
Choose a rhythm you can keep for three months, even a low one: one long video a week and two Shorts are more sustainable than five releases you skip. Consistency is a recurring tip in industry guides, which come from vendors and so are not measurements; the only measurement that counts is yours, on your channel. To organize the work, see content repurposing for creators.
Which metrics are worth watching?
Three are enough: retention after the first 30 seconds, views from people who don't follow you and subscribers gained per video. YouTube defines intro retention as the share of viewers still watching after the first 30 seconds.
The key-moments report, according to the official help page, needs a video of at least 60 seconds and at least 100 views, and data arrives after one or two days. We found no official value for a “good click-through rate”: the YouTube pages we read give no benchmark. So compare your own thumbnails with each other and see how retention changes after the click.
A sheet to assess your first ten videos
After each video, fill in a row with four items: the title, retention after 30 seconds, views from non-subscribers after a week and subscribers gained. Add a column with one sentence: “What I would change”. After ten rows you see the patterns: which openings hold people, which titles bring new viewers, which topics make them subscribe.
Don't compare your videos with those of bigger channels: compare them with your own earlier ones. The useful signal is the direction, not the absolute value. If retention rises when you shorten the intro, shorten it more; if views from non-subscribers grow with a certain kind of title, use it again. The sheet is also a good way not to be swayed by a single viral video, which can be down to chance.
When can you monetize a channel?
The Partner Program needs at least 1,000 subscribers and 4,000 public watch hours in the last 12 months, or 10 million Shorts views in the last 90 days. Review usually takes about a month, and you can reapply if you are rejected.
- Subscribers
- 1,000
- Public hours in 12 months (or)
- 4,000
- Million Shorts views in 90 days
- 10
The page warns that from 1 February 2027 the program introduces updates and that creators must accept the new terms in YouTube Studio by 31 January 2027. It gives no new thresholds: check it before you count on ad revenue. To avoid depending on one source, read how creators make money and creator newsletters.
How do you organize the first 90 days?
Split them into four phases: set up, publish and observe, correct, and decide what to double down on. Each phase has one job and one question to answer with your own data.
- Weeks 1-2: set up. Write the promise, pick three topics and prepare the first two videos.
- Weeks 3-6: publish and observe. Release consistently and note the three numbers each week (retention at 30 seconds, views from non-subscribers, subscribers gained).
- Weeks 7-10: correct. See which videos hold viewers best and change one thing at a time: the opening, the title or the length.
- Weeks 11-13: decide. Double down on the format that worked, drop the others and write down what you learned.
The most frequent mistakes are three: imitating formats that don't suit you, changing everything every week and ignoring retention. If you use AI tools for captions or ideas, remember the disclosure rules we cover in AI for creators.
An example of a typical week
A sustainable week might be: Monday planning of a long video (outline and three points to make), Tuesday and Wednesday recording, Thursday editing, Friday publishing plus two vertical clips taken from the same video, Saturday reading the three numbers. Weekend work is not compulsory: if you skip a week, resume without catching up.
The point is that creative work and analysis sit in different moments. Looking at the numbers right after publishing leads to changing your mind out of anxiety; looking once a week, with a cool head, leads to better decisions. Keep a single sheet with the date, the title and the three numbers, and reread it every four weeks.
What to do if the channel hasn't taken off after three months
Before you change everything, recheck three things: the promise (is it clear to someone arriving for the first time?), the opening of your videos (what happens in the first thirty seconds?) and the topic (do you have enough real questions to carry on?). Often the problem is a single one, and fixing it is worth more than adding new ideas.
Ask five people from your ideal audience for an opinion: they watch two of your videos and tell you where they got bored. If retention is low after 30 seconds, rework the start; if views from non-subscribers are few, rework the title and promise. If after another three months of corrections the numbers do not move, you may need to change topic or format: it is a legitimate decision, to be taken with data and not with frustration.
Key terms. Retention is the share of viewers who keep watching a video after a given moment. A Short is a vertical or square video up to three minutes long. The Partner Program is YouTube's program that lets you earn from ads and other features.
How we checked this. We found in the official recommendations page the signals listed, but no weights or figures. We found in the retention help page the requirements of 60 seconds and 100 views. We found no official benchmark for click-through rate. Contact: hi@heytea.app. Editorial note: we fact-check every figure against the sources listed below and correct mistakes openly.
Frequently asked questions
How do you grow on YouTube with no subscribers?
You start with a clear promise for a specific audience, publish consistently and watch a few numbers: how many viewers stay past the first 30 seconds, how many views come from people who don't follow you and how many subscribers you gain. YouTube publishes no formulas: it only names the signals it considers, such as viewing behaviour, subscriptions and viewer feedback.
Do Shorts help a channel grow?
They bring visibility more than loyalty. In Metricool's study, a tools vendor, the Shorts feed generates 61.54% of the sample's views, but subscribers remain a separate source (11.31%) and the page does not say how many Shorts viewers become subscribers. Use them to be discovered and link them to a long video.
How many videos should you post a week?
There is no official number. In Metricool's sample, channels posting 2 to 4 Shorts a week got the best results per post, but it is a vendor's figure and causation is not shown. Choose a rhythm you can keep for three months: consistency matters more than intensity.
When can you monetize a YouTube channel?
The Partner Program needs at least 1,000 subscribers and 4,000 public watch hours in the last 12 months, or 10 million Shorts views in 90 days. Review usually takes about a month. From 1 February 2027 the program changes its terms: check the help page before you plan around ad revenue.
What is a good click-through rate for thumbnails?
We found no official value: the YouTube pages we read give no benchmark. Compare your own thumbnails with each other, on the same topic, and watch how retention changes after the click. A high click-through rate with viewers who leave at once is a sign of a promise the video does not keep.
Sources
- YouTube, “How YouTube Works”, recommendations. Official page, undated. Read on 16 October 2026. youtube.com
- YouTube Help, YouTube Partner Program: requirements and terms updates. Read on 16 October 2026. support.google.com
- YouTube Help, “Measure key moments for audience retention”. Read on 16 October 2026. support.google.com
- Metricool, “Algoritmo YouTube Shorts 2026”, 31 August 2026. A vendor source, its own study of 799,718 videos; period and methodology not stated on the page. Read on 16 October 2026. metricool.com