Loading...
Loading...
This video is about a personal network system tour and mentions products and tools such as a 4G wireless LTE modem, fiber internet, LabRax, TP-Link router, Omada system, Pi-Hole, Nginx, Technitium, and a C100 controller.
here is my network rack, if you were interested. I use TP link Omada as the controller, and this gives me the ability to create and manage all the vlans I need. pihole, nginx and technitium manage all my dns. then I have a switch to the 3 accespoints for WiFi and then my second switch in the server rack for all the other lab stuff.
Performance Category
Average
Score
3.0/5
Shares: 1/5
Comments: 3/5
Retention: 5/5
Views: 3/5
Likes: 4/5
Followers: 3/5
Script: 2.4/5
Total Views
4361
Likes
203
Shares
1
Comments
11
Duration
0m 59s
For You
4,104
94.1% of views
Personal Profile
113
2.6% of views
Follow
105
2.4% of views
Search
26
0.6% of views
Others
9
0.2% of views
Direct Message
4
0.1% of views
Sound
0
0.0% of views
Views
Likes
Shares
Comments
For You Traffic
Profile Traffic
Search Traffic
Non-Followers
43.0%
1,875 views
Followers
57.0%
2,486 views
8.8% of followers reached
New Followers
5
Performance vs Median
Transcript Available
Perceived Value
48/100
Compared to Average
Average
"A while ago, I promised you that I would give you a bit of a ne- a tour of my network system."
"I am running two internet connections. We have a 4G wireless LTE modem and a fiber."
"So these come into the TP-Link router, and I have this as a load balanced, so if one drops, the other one takes over."
"Let me know if you have any questions. Happy to answer."
The opening does not make it obvious why a viewer should keep watching. It starts with backlog context instead of a compelling reason: retention falls from 100% at 0s to 83% at 1s and 65% at 2s, then to 55% at 3s and 47% at 4s. That early drop strongly suggests the first lines did not create urgency, promise, or curiosity. The like distribution also supports this: 59% of first-15s likes happen at 0s, then only 8% at 1s and 4% at 2s, meaning the opening did not continue earning interest after the first frame.
Replace the intro with a concrete benefit in the first sentence. For this script, try: "Here’s my failover home network setup: dual internet, automatic backup if one dies, and local DNS with Pi-hole." That immediately tells viewers what they will see and why it matters.
"“A while ago, I promised you that I would give you a bit of a ne- a tour of my network system, uh, and”"
The script eventually signals that this is for networking or homelab viewers, but it does not do so quickly enough. The audience signal only becomes clearer around 5s to 10s when you mention 'two internet connections' and specific network hardware. By then retention has already dropped from 100% to 42% by 5s and 39% by 6s, meaning a majority of viewers left before the content identified who it was for.
Name the audience immediately. Open with language like: "If you're into homelab or home network redundancy, here's my dual-WAN setup." That filters in the right viewer in the first second.
"“I am running two internet connections. We have a 4G wireless”"
There is no clear problem stated upfront. The script is mostly a tour, but it does not anchor itself in a concrete pain such as internet reliability, ad blocking, DNS control, or failover. Without a defined tension, the content feels observational rather than need-driven. This likely contributed to the sharp early retention decline from 100% to 55% by 3s and to 34% by 8s.
Frame the setup around a real problem: reliability, speed, privacy, or uptime. Example: "I got tired of internet outages killing my work, so I built a dual-connection network with local DNS and automatic failover."
"“so let's have a look. I am running two internet connections.”"
The viewer is not told what they will gain by watching to the end. 'Let's have a look' promises only a casual walkthrough, not an outcome, lesson, or reveal. Retention dropping to 42% by 5s and only 32% by 10s suggests viewers did not see enough future value in staying. Shares are also extremely low at 1 total share, which often correlates with weak takeaway clarity.
State the payoff explicitly: what will the viewer learn, copy, or avoid? Example: "In under a minute, I’ll show you how I set up dual internet failover, local DNS, and centralized control for my home network."
"“I never did, so let's have a look.”"
Once the setup starts, the script contains a fair amount of actual information: dual internet, TP-Link router, load balancing, Omada controller, Pi-hole, Nginx, and Technitium. That supports a mid-level score. However, the value arrives after a slow opening and is delivered as a list rather than a sequence of insights. Retention flattens somewhat after the initial collapse, from 39% at 6s to 29% at 14s, which suggests the people who stayed found some value, but not enough to re-accelerate engagement.
Compress filler and attach each detail to a purpose. Instead of listing components, use a format like: 'Two connections for redundancy, TP-Link for failover, Omada for control, Pi-hole for filtering, Technitium for local DNS.'
"“This is running Pi-Hole, as well as Nginx and Technitium.”"
The script uses strong concrete nouns and named tools: 4G LTE modem, fiber, TP-Link router, Omada controller, C100, one-gig switch, Pi-Hole, Nginx, Technitium. That gives the content technical texture and makes it credible for a niche audience. This is one of the stronger areas of the script. The modest like count of 203 on 4,361 views suggests some viewers did appreciate the specifics, even if the broader audience did not stay long enough to hear all of them.
Keep the specificity, but front-load the most meaningful specifics earlier and tie them to outcomes. Example: 'Dual-WAN TP-Link failover plus Pi-hole and Technitium for fully local DNS.'
"“This is an Omada system, so everything is controlled by the Omada controller, the C100”"
The script sounds like it comes from real hands-on experience. You mention exact hardware, architecture, and operational choices like load balancing and local upstream DNS. Those are credible details that are hard to fake. The low share count and average retention do not undermine credibility; they suggest packaging issues more than expertise issues.
Add one practical trade-off or reason behind a decision to strengthen expertise even more. For example: 'I use Technitium upstream so I keep DNS local and avoid relying on remote providers.'
"“I have this as a load balanced, so if one drops, the other one takes over.”"
The transcript feels human and unscripted, with minor hesitations and direct commentary like 'not so great here.' That helps it sound real rather than polished marketing. This likely helped preserve some trust among the viewers who stayed, even as the opening lost broad attention. Comments at 11 also suggest a small but real niche audience connection.
Preserve the natural tone, but make the first line more intentional. You can still sound authentic while being clearer: 'I kept saying I'd show my network setup, so here it is—built for redundancy, local DNS, and easy control.'
"“we have, mm, not so great here. This is a 3D printed LabRax.”"
The spoken phrasing is somewhat hard to parse in real time because it includes interruptions, restarts, and chained references without clear signposting. The first sentence is especially messy, and that likely contributed to the steep retention decline from 100% to 65% by 2s. Even later, the script stacks component names quickly, which may overwhelm non-expert viewers.
Use short, single-purpose lines. Break the explanation into clear chunks: 'Internet in. Router. Controller. DNS box.' You can also pair each object with one job to reduce processing load.
"“A while ago, I promised you that I would give you a bit of a ne- a tour of my network system, uh, and”"
The script moves physically through the setup, but not narratively through hook, problem, method, proof, and takeaway. It feels like a room tour rather than a structured explanation, so each section does not build enough momentum. The retention trend supports this: after the early crash, it declines steadily from 42% at 5s to 29% at 14s without any visible re-engagement spike. That suggests the video did not create new layers of payoff as it progressed.
Rebuild the sequence as: problem, solution overview, component breakdown, why each matters, final takeaway. Example: 'My goal was uninterrupted internet. So I use dual WAN, then Omada for control, then Pi-hole plus Technitium for local DNS.' End with a summary benefit instead of only an invitation for questions.
"“That is the network. Let me know if you have any questions.”"
lirkangelas i understand u have static ip? also from provider u got bridge mode and point to router? after router going to patch that will split on each vlan? omada is taking as only AP?
ahh, nope. no static IP I'm behind a cgnat so that's not possible. I use tunnels when I need to expose services publically.
the controller is a ui manager, the vlans are all managed through there, so it's vlans on the router and managed though the controller.
JonWhat’s the wifi setup?
I have the tplink WiFi 7 access points, the ones that are POE.
stevoau7Can we see ur sever and storage? Would be interested to see what devices ur using
yeah, I'll do a second one tomorrow for you.
stevoau7Thx so much - watched it and it was great. I wish I got into home labs earlier cause now too expensive
second hand, facebook marketplace, and office disposal stores, all you need and no where near the cost. don't buy new.
TimmmeeehhhDi you have a video about Technitium? i run adguard but don’t run my own dns for upstream, is it worth it?
its very low effort, I don't have a video on it but I'll make one for you. ;)
Total viewers and likes aligned with spoken words.