The "What Would I Want to Watch?" Filter
Run every content and business decision through your own taste as the audience
- Difficulty
- Easy
- Time to result
- ~weeks to results
- Steps
- 4
- Confidence
- 82%
The filter collapses dozens of creator decisions into a single recurring question: would we want to watch this, and would we want to subscribe to this channel? Brownlee runs it across the whole operation, from which video to make, to whether to take a sponsor, to whether to plug merch. Because the maker's taste stands in for the audience's, the filter keeps output coherent and prevents the slow slide into view-chasing or ad-stuffing that would technically make more money but produce a channel the maker himself wouldn't follow. It also solves consistency: when you only make things you genuinely want to watch, sustaining output for a decade stops feeling like a grind. The filter is deliberately blunt, which is what makes it fast enough to apply to every decision.
Origin
Extracted from Deep Dive with Ali Abdaal. Brownlee describes it as the overarching goal he keeps in the back of his head for everything MKBHD does, from choosing videos to hand-picking rare sponsors.
Core principles
- 01You are a proxy for your ideal audience, so your own taste is a usable signal
- 02The same question governs videos, sponsors, formats and hiring
- 03Making what you'd want to watch beats chasing what will perform
- 04Consistency comes from enjoying the thing you produce
- 05Say no to anything you wouldn't subscribe to yourself
How to run it
- 1
State the decision as a viewing choice
Reframe any pending choice, whether a video idea, a sponsor, or a format, as: is this something I would want to watch?
Pro tip Add the second half, 'would I want to subscribe to a channel that does this,' to catch decisions that look fine once but corrode the brand at scale.
- 2
Apply it beyond content
Use the same question for sponsorships, merch plugs and expansion decisions, not only for individual pieces of content.
Watch out A single sponsored video passes easily; imagine a whole channel of them to see whether it really clears the filter.
- 3
Say the hard no
Decline opportunities that fail the filter even when they are lucrative or on-trend.
Pro tip Track how many no's you say; a healthy filter produces a full inbox of declined offers.
- 4
Break ties with genuine interest
When several ideas pass, pick the one you are most personally excited about rather than the one you predict will perform best.
In the wild
Brownlee notes he could run a sponsored segment on every single video for the rest of time and make far more money. He runs each offer through the filter instead: would he subscribe to a channel that did that? The answer is no, so the vast majority of sponsor emails become polite declines, and only win-win products he'd share anyway make the cut.
→ A channel with unusually few sponsorships and preserved audience trust.
Common mistakes
Using view predictions as the filter
Swapping 'would I watch this' for 'will this get views' gradually turns the channel into something optimized for a metric rather than for the maker's own standard.
Applying it only to content
Creators forget to run sponsors, merch and expansion through the same question, which is where most brand erosion actually happens.
Is it for you?
Best for
Creators and founders whose personal taste closely overlaps with their target audience.
Not ideal for
Products aimed at an audience whose preferences are nothing like your own.
From the transcript
“we want to make what we'd want to watch”
“We wouldn't want to subscribe to that channel, but we could do it if we wanted to make a lot of money.”
From the episode
MKBHD: The Story To 16 Million & The World Of YouTube
MKBHD