Dave Lu (@DaveLu_creator) profile picture

Dave Lu

@DaveLu_creator

Faceless YouTube + TikTok. Hooks, formats, AI video, growth. I work with ShortsFast ↓

shortsfast.com
Joined September 2026
  • Tweets92
  • Following21
  • Followers2
You can launch a new YouTube channel with your old one as the co-host. @NickBYT_ used YouTube Collaborations to pair each of his first four uploads with his main channel. He reports 300K+ views across those four.
1
Search "grocery haul" before you script a TikTok for a budgeting app. Find a haul that opens on the receipt. In your version, show the total first, then reveal which item blew the budget. That's my spin on @tuanle's research trick.
2
"Who gets the money from a $5 coffee?" is a whole faceless episode. Follow the money from the register back to the farm. That's the "money trail" format hiding in @successfaceless's list of 50.
6
Photography and pickleball sharing a YouTube title formula is wild. "80% of [skill] in [time]." @FaradayYouTube says 24 of 40 examples beat their channel's usual views by 5x+. Six underperformed. I'd study both groups.
5
30M+ views from the first three videos. @danvsI says that was Simplihow's launch two years ago. Then he casually mentions posting for brands on TikTok Shop as a side project. Okay, I want the side-project breakdown.
4
Damn. @keneth_ai made an animated romance with Seedance 2.5 in CapCut. Someone in the replies says the guy gave her a ring too soon. We're giving dating advice to AI characters now.
6
I work with ShortsFast. Before batching faceless videos, finish one script through its last line. Rewriting the ending during production leaves you sorting through versions of an idea you still haven't settled.
1
8
Try a script about keeping a houseplant alive. If the opening promises to explain yellow leaves, the ending needs to answer that question. 'Follow for more plant tips' can come after the answer; it can't stand in for it.
1
5
Mark that script ready only after the opening and ending agree. Then make the video and review what the finished version actually says. For AI video creation, ShortsFast is linked in my bio.
2
A useful YouTube tutorial can answer the viewer's entire question. Once the problem is fixed, subscribing may never cross their mind. Before changing your editing, check what the channel offers that same person next.
1
3
Say your video helps someone free up space on their phone. A follow-up about backing up their photos gives that viewer another reason to watch. An unrelated gadget unboxing asks them to start caring about a different subject.
1
3
I'd make that next video easy to find from the first one, then watch whether people take the next step. If they leave after getting the answer, the original video may have done its job. That alone doesn't tell you the channel has a reason to be followed.
1
A faceless TikTok Shop video can show a dress without showing your face. I'd be cautious about using an AI try-on as proof of how that dress fits. The buyer needs to know which details came from the actual garment.
1
7
Film the fabric in your hands or show the real hem and seams. If you're describing stretch, demonstrate it with the item. A generated mirror selfie might suggest an outfit, but I wouldn't use it to answer a shopper's sizing question.
1
5
Before making a batch, read the questions shoppers keep asking about that product. Pick one you can answer with real footage. If the question is whether the fabric is see-through, a polished avatar gives me very little to go on.
1
Your YouTube watch hours don't all reset when your channel turns a year old. For the watch-hour eligibility route, YouTube looks at the last 12 months. Older hours fall outside that window as newer hours come in.
1
8
A public long-form video can keep earning new qualifying hours long after you upload it. A tutorial posted last year can still contribute when someone watches it this week. The upload date alone doesn't make that watch time expire.
1
4
Use the Earn tab in YouTube Studio for eligibility progress. Shorts Feed watch time doesn't count toward the 4,000-hour route, so your total watch time in Analytics can tell a different story. I'd check which counter you're trying to grow.
1
I work with ShortsFast. Before dubbing a faceless YouTube video into ten languages, I'd get one language checked by someone fluent. A convincing voice wouldn't be enough for me to approve it.
1
3
For a history video, have the reviewer check a passage with names and a joke that depends on the wording. Ask them to explain it back to you. 'The voice sounds natural' doesn't tell you whether the story survived translation.
1
3
Then use the audio track filter in YouTube Analytics to see how that version performs. If viewers leave at a passage the reviewer flagged, revisit it before expanding. I'd keep the original script beside the translation.
Load more