Mel Robbins' YouTube Strategy: What All 1,503 Videos Reveal

Those two videos are six months apart on the same channel, shot on the same equipment by the same team. One has 727 times the views of the other.

The difference is not production. It is what each thumbnail offers, and the pattern behind it runs through all 1,503 videos.

The formulas

Eight came out of the analysis. Each was tested against her own era, then re-tested with the top 1% of videos deleted and on the older and newer halves of the library separately. The strength note on each one is what survived.

• Long-form titles: name the credential. Titles carrying a doctor, professor, neuroscientist or surgeon run 1.86x the rest. 47 videos, and it held on every cut.
• Long-form: go longer. Since 2023 the bands run 0.19x, 0.81x, 1.20x, 1.77x as length rises. No exception at any band.
• Long-form titles: never write in the first person. "I", "my" and "me" cost about 28%, and that held on every cut.
• Titles: use "how to". Worth 1.33x on her long-form and 1.45x on her Shorts.
• Shorts titles: put your own name in. Worth 1.39x across 807 of her Shorts.
• Never publish housekeeping. Q&A, live and announcement titles run 0.45x, and 0.11x since 2023.
• A colon still helps, but less each year. 1.35x across the library, 1.13x on the newer half.
• Question marks cost a little. 0.91x across the library, 0.72x since 2023.

First, the thing everyone asks

The first objection to any of this is that the views were bought. The like rates say otherwise. Bought views watch and do not press like, so the like rate falls away, and hers does the opposite. Her six biggest videos run like rates between 2.3% and 4.1%, and comment rates between 0.07% and 0.21%, which is her channel's own normal at that reach.

Her oldest hit is the strictest test, because it is the one with nothing to hide behind. "The Secret to Stopping Fear and Anxiety", published in 2017, has 5,438,729 views and 170,260 likes. That is 3.13%, on a video nine years old.

Then, what it isn't

It isn't the old catalogue carrying her. Her biggest 2017 video has 5,438,729 views. A 2024 upload has 5,052,269 and a 2025 upload has 5,000,020. The new work matches the legacy work.

It isn't Shorts. They are 60.5% of her library and 37.4% of her views.

It isn't a quiet year. She published 421 videos in the twelve months to 20 August 2026 and gained 1,670,000 subscribers, reaching 6,050,000.

Four things are left.

1. She stopped being the expert and started booking them

Mel Robbins is not a doctor. Since the podcast took over her channel, most of her best videos are someone who is.

Long-form titles naming a medical or academic credential run 1.86x the rest of her long-form. That is 47 videos, and it is the sturdiest number in this article: 1.86x with the top 1% of videos deleted, 1.59x on the older half of the library, 2.01x on the newer half, 1.90x across the podcast era alone.

The top of her library reads like a directory: a Stanford neurosurgeon, a Harvard addiction specialist, a Mayo Clinic cancer doctor, the #1 menopause doctor, Harvard Business School's #1 professor, a trial lawyer, an exercise physiologist.

Her name is still on the thumbnail. The credential is what the title sells.

The Takeout: If you are the only expert your channel ever features, you have capped it at the size of your own authority. Booking someone with a credential you do not have is the cheapest way to raise that ceiling.

2. She went the wrong way on length, and won

The woman who made her name on a five-second rule now publishes hour-long episodes. In 2017 her median long-form video ran 6.7 minutes. This year it runs 68.2 minutes, and 99% of her long-form is over 35 minutes.

The obvious reading is that she got away with it. The data says the length is doing work.

In raw terms the same pattern holds: a median of 34,552 views under 30 minutes, against 684,404 over 90.

I cannot prove the mechanism from outside. Retention and traffic sources are visible only to the account owner. The honest version is that her long videos are not long recordings of short ideas. They are one guest, one subject, and a chapter list that reads like a contents page.

The Takeout: Before you cut a video down, check whether the subject is worth an hour. A thin idea stretched to sixty minutes fails, and so does a deep one cut to fifteen.

3. The Shorts cost her almost nothing

She publishes roughly three Shorts for every long-form video, and the gap between what they are worth has widened every year.

In 2022 her median long-form video was worth 1.3 Shorts. This year it is worth 7.3. She has not cut the Shorts, and the share of her uploads that are Shorts has risen to 77%.

The Shorts are cut from episodes she is already recording, so they cost her almost nothing, and they are where a stranger meets her first. The episodes are where the audience is built.

The Takeout: Judge Shorts on what they cost, not on what they return. If yours are cut from long-form you were making anyway, a falling Shorts median is not a problem. If they are made from scratch, it is.

4. What kills her videos is asking for something

Sort her long-form library by performance against its own era and the bottom is not random. It is the housekeeping: Q&As, live sessions, announcements and the videos titled after her own news.

Her five worst-scoring long-form videos	Views	Score
Answering Your Questions... Q&A	6,951	0.04x
I Never Do This... Late Night Q&A!	8,165	0.05x
The Path Behind You Is No Longer An Option.	13,791	0.08x
There is something I need to tell you	17,871	0.09x
You Are Meant For More. Watch This	16,929	0.10x

As a group, housekeeping titles run 0.45x the rest of her long-form, and 0.11x since 2023. It is a small set, 16 videos, so treat the exact figure as a direction rather than a law. The direction is not subtle.

Her worst-scoring long-form is the most instructive video on the channel. It is a live Q&A published in April 2024, and about a third of the way in it becomes a pitch for her six-month coaching programme, with a closing deadline. It has 6,951 views on a channel with six million subscribers.

The same channel, in the same month, was publishing hour-long episodes doing hundreds of thousands.

The Takeout: The videos where you ask are the ones nobody watches, so put the ask inside the videos where you give. A launch video published on its own reaches a fraction of your normal audience.

What the transcripts show

I pulled the openings of her best and worst episodes.

Her third-biggest long-form, "The Best Financial Advice You'll Ever Hear", opens at 0:00 on her guest, not on her:

"What's the number one thing that keeps people broke?" "It's ignorance. It's not a lack of intelligence. It's ignorance."

Mel's first words arrive after that, and they are a promise with a credential attached: the best financial advice you'll ever hear, from the man who changed how she thinks about money, ten million copies sold. The whole apparatus is built before the first minute is out.

The Q&A opens like this:

"hey oh my gosh this worked I'm so excited to be here with you um I have been wanting to get on live and start giving you uh coaching and answering your questions live for a long long time"

Two minutes in, she is still waiting for people to arrive in the chat.

The Takeout: Give the first twenty seconds to the viewer's problem or to someone with a credential. Anything you say about your own excitement is spent from the same budget.

The search layer

Her strongest new subject is money, and it is also the most valuable ground she stands on. Since 2024, money and career titles run 2.66x the rest of her long-form. That is only 12 videos, so on its own it would be a hint. The raw numbers back it: "The Best Financial Advice You'll Ever Hear" has 5,000,020 views, "5 Money Rules" has 3,914,828, and "Why You're Broke" has 1,888,748.

Her audience is mostly American, so these are United States Google Ads figures, in US dollars, pulled on 20 August 2026. Volume counts searchers. Cost per click is what advertisers will pay for one of them.

Term	Searches/mo	CPC	Advertiser market
narcissist	673,000	$0.32	$215,360
mel robbins	201,000	$0.40	$80,400
menopause symptoms	110,000	$2.25	$247,500
personal finance	22,200	$12.09	$268,398
the 5 second rule	18,100	$1.38	$24,978
how to stop overthinking	14,800	$2.15	$31,820
how to stop procrastinating	9,900	$1.99	$19,701
how to invest money	8,100	$7.70	$62,370
how to manifest	8,100	$0.86	$6,966
how to build confidence	5,400	$1.32	$7,128

Read the first and fourth rows together. Narcissism brings thirty times more searchers than personal finance and advertisers pay a thirty-eighth as much per click, so the smaller subject is worth more per month to the people buying the traffic.

That is the pattern across her whole subject list. The topics that bring her the most people sit at the cheap end, and the one advertisers pay most for is the one she started last.

The same check settles a question the analysis raised. Her signature idea has not faded: "the 5 second rule" still takes 18,100 US searches a month and her own name takes 201,000. In the UK the same phrase takes 720. Two markets, one phrase, a 25-fold gap, which is why the market has to be named whenever a search figure is quoted.

The Takeout: Volume counts searchers and cost per click prices them. Check both before you commit a year of content to a subject, because the biggest lane in your niche is often the cheapest one.

One rule that did not survive

Putting a word in capitals looked like a rule. Across the whole library, long-form titles with an ALL-CAPS word run 1.32x the rest, which is the sort of number that ends up in a listicle.

Then I cut it the other ways. On the older half of her library it is 1.59x. On the newer half it is 0.70x. Across the podcast era it is 0.87x. It marks the era when she was making short motivational videos, and those did well for reasons unrelated to capital letters.

Every other formula above cleared that test. This one is published because it did not.

The uncomfortable version

She gives away more properly researched health and money content than most people sell, she books guests who are better qualified than she is and lets them talk, and she has added 1,670,000 subscribers in a year while doing it.

The uncomfortable part is what she gave up to do it. She had a technique with her name on it, and the channel that works now is built on other people's credentials, in subjects she is not the authority on. Her median video is ten times longer than the idea that made her.

She did not defend the five-second idea. She became the person who books the expert, and the long-form median went from 105,115 to 475,383 across the two years that followed.

Method and limits

Public data only. The full library of 1,503 public videos was exported on 20 August 2026, with subscriber and view history from tracked daily counts over the twelve months to that date, and transcripts from auto-generated captions. Views, publish dates and durations come from that export; likes and comments in the export are rounded to two significant figures, so every like rate and comment rate quoted above was re-pulled at full precision from the YouTube API on the same day rather than taken from the file. The export's view total sits 2.0% below the total the channel reports, which is the usual gap between a library export and a lifetime counter.

Comparisons are era-normalised: each video is scored against the median of its own calendar half-year, in its own format, and only videos at least 90 days old are scored. That covers 559 long-form videos and 819 Shorts. Shorts are defined as three minutes or less. Every title rule was re-tested with the top 1% of videos by views deleted, on the older and newer halves of the library separately, and on the podcast era alone; where a rule moved, the article says so.

All view figures here predate 24 August 2026, when YouTube changed what counts as a view on long-form, so nothing above spans that boundary. Search figures come from Google Ads for the United States, her main market, and are in US dollars, which is the currency that source reports. An earlier draft used United Kingdom figures, which understated "the 5 second rule" by 25 times; the market a figure is pulled for is named above for that reason. Every term quoted was also checked for meaning rather than assumed: the YouTube results page and autocomplete for "the 5 second rule" return her book, her talk and her own uploads rather than the dropped-food idiom, and the Google results page returns her book and a video pack of her clips, so the volume belongs to her idea. Google Trends for the United Kingdom over the 112 weeks to 20 August 2026 returns data for that phrase in only 11 of them, which is why the UK figure is small rather than declining. Retention, click-through rate, traffic sources and revenue are visible only to the account owner and are not estimated anywhere above. YouTube reports subscriber counts to three significant figures, so the twelve-month deltas are approximate. Video counts are what the channel still shows today, which is a floor rather than a total, since deleted uploads are invisible from outside.

Analysis and write-up: Stewart Read.