YouTube strategy Q&A: research, packaging and viewer satisfaction.

From 0:00 →

Long-form: do the research before you record

Short-form lets you test hooks cheaply; a long video has to be made properly to be a fair test. Jake says the research should come first so the idea is a calculated bet, and that his team often does one video a week for clients so they can diagnose a weak result.

“But you can't properly test a YouTube video idea if you don't do the video properly.”Jake, 0:15

From 1:30 →

Borrow from adjacent niches, not your own

Jake’s reasoning is that copying what works inside your niche will not beat it. Instead look at what resonates with viewers who think the way yours do, in nearby niches.

“You don't want to copy what's working within your niche, otherwise you're never going to beat the market.”Jake, 1:51

From 2:42 →

An example: a speedrunning format

The team saw a speedrunning-style format work in the drop-shipping niche, and used that as evidence it could resonate with a business audience, then adapted it to cold calling. Jake describes the result as hundreds of thousands of dollars of revenue for the client, a figure he states in the talk.

From 4:29 →

Views first, then find the bottleneck

For a channel starting from zero, Jake treats views as the primary measure because too few views gives too little data. Once traffic is enough, tracking shows whether the loss is in the video, the call to action or the funnel.

“if you don't have enough views, then you don't have a large enough data set to actually make decisions.”Jake, 5:12

From 6:46 →

Persona: say who the video is for

Beyond the hook, proof, promise and plan, Jake adds a fourth P: persona. Name the core market first so they know it is for them, and match the proof to what the viewer clicked for.

From 10:09 →

How viewer satisfaction is read

Jake describes it as a mix: the like-to-view ratio, the share of positive comments, comment sentiment, watch time and how long people stay on the channel afterwards. He says early positive comments can be a leading indicator, provided the comments are organic.

From 12:29 →

Diagnosing an underperforming video

Analytics are signal, and the diagnosis comes from digging further, for example an on-screen habit that makes a video feel read from a script, or a dip in the retention graph where a hard call to action appeared.

From 15:37 →

When to judge a video

Newer channels can take a month or more to gain traction. Check when the video was posted, how many comments it has and whether the views are organic rather than paid before treating it as evidence.