Does learning with your own content actually improve long-term vocabulary retention?

I’ve been thinking a lot about why vocabulary often fails to stick long-term, even when learners are consistent and motivated.

One idea that keeps coming up is learning through your own content — words and sentences you personally needed, looked up, or wanted to say — instead of predefined courses or generic decks.

I’m curious how people here see this from real experience, both as learners and as builders.

Specifically:

• Does using your own words and sentences actually improve long-term retention compared to pre-built content?
• How important is audio in this process — especially listening to your own content repeatedly, almost like a personal podcast?
• Have you seen cases where listening helped vocabulary stick more than active drills alone?

From a product perspective, I’m also interested in the trade-off:

Many apps optimize heavily for retention mechanics (streaks, gamification, pressure).
But those don’t always translate into real learning.

• Is it possible to design a language learning tool that prioritizes actual retention without relying on aggressive engagement tricks?
• Where have you seen this balance work — or fail?

I’d really value:

• Learners’ experiences with what actually worked after motivation faded
• Builders’ lessons from products that tried to respect attention while still staying viable

Thanks — genuinely trying to understand what leads to durable learning, not just usage metrics.