Alex Lukianov retweetledi
Alex Lukianov
920 posts

Alex Lukianov
@LigoAlexandr
Entrepreneur. Growth Hacker. Builder.
Katılım Aralık 2014
181 Takip Edilen86 Takipçiler

Everyone talks about "TikTok Organic" like it's easy money.
"Just post 3 times a day and go viral."
But when you actually try it?
0 views. 200 views jail. Shadowbans.
We burned dozens of accounts and over $10,000 figuring this out.
We thought our content was the problem.
It wasn't. It was our Infrastructure.
We were making invisible mistakes that killed our reach before we even started:
❌ Using "dirty" Apple IDs
❌ Leaked IPs
❌ Rushing the warmup phase
❌ Scaling too early
❌ Copying formats too late
Most guides fail because they ignore the tech stack.
They teach you to be a creator, not an operator.
Once we fixed the backend, everything changed.
Now we generate 500k-1 Million organic views/month.
I documented our entire internal SOP.
~50 pages of pure operator logic. No fluff.
It covers:
→ Infra (Device & Network)
→ Warmups (The exact protocol before posting)
→ Testing (Validating account health)
→ Decay & Recovery (How to fix view drops)
→ Scaling (Safe expansion SOPs)
→ Kill Rules (Knowing exactly when to abandon an account)
I'm giving the full manual away for free.
Want it?
1. Like this post
2. Comment “GUIDE”
3. Follow me (so I can DM you the link)

English

@LigoAlexandr Hi Alex. Really love the content you are putting out. These are real bangers. Please message me I would like to do business with you.
English

🧵 TikTok Farm Playbook — Part 5/20: Warmup Phase 0-1 (why your first 5 days determine everything)
⬆️ Previously: Mobile proxies > residential > VPN. TikTok only checks ASN + GEO. Fraud scores are useless.
This is the step that separates farms that print money from farms that burn accounts.
90% of people create an account and post within hours. Every single one of those accounts dies under 300 views. Every. Single. One.
TikTok builds a behavioral model of your account BEFORE it distributes your content. It needs to understand who you "are" - what niche, what audience, what engagement patterns.
Phase 0: Lurker (Days 1-2)
- Scroll FYP 5 min per account, 2x/day
- Watch videos to COMPLETION
- Like only in your exact niche
- Follow 5-10 relevant accounts
- DO NOT POST ANYTHING
Phase 1: Training (Days 3-5)
- Search keywords in your niche
- Filter by "last 3 months"
- Like + comment under relevant videos
- Leave comments like "this is interesting" or "never thought about it like this"
- Avoid emojis and questions early
- Follow max 5 new accounts per day in niche
Goal: your FYP should be 70%+ niche-aligned with mostly US-based creators. When you see that - you're ready to post. Not before.
We tested skip-warmup vs 5-day warmup across 50+ accounts. The warmed ones had 3-4x higher survival rates.
⬇️ Next: Your first posts and the 700-view health test that tells you if your account is alive or dead
🔁 Repost if useful. Follow for the full series.
English

The thing that kills AI influencer accounts fastest isn't people getting tired of the character - it's the platform catching the pattern. SynthID and C2PA metadata flags are getting better at detecting AI-generated images, and the telltale sign is an initial view spike followed by a hard drop once the classifier kicks in. Slideshows with mixed real + AI frames actually dodge this better than full AI video right now. Worth keeping that as a fallback format when the detection tightens up
English

I built an AI influencer that posts content 7 days a week.
She always submits content on time, never complains about doing a reshoot, and slowly pushes qualified traffic to my products.
Here's how the system works:
→ Generate a hyper-realistic AI character in Calico AI
→ Lock in facial consistency across every single shot
→ Build a content calendar of lifestyle scenarios — morning routines, errands, gym, late night spots
→ Render each scene with different outfits, lighting, and settings
→ The audience sees the same "person" showing up again and again
That's how you build brand recall without ever hiring a creator.
Today, this tactic crushes.
Tomorrow, it will be worn out and people will get sick of it.
These windows of opportunity always only last so long...
Comment "INFLUENCE" and I'll send you the exact system (must be following so I can dm you!).
English

The cost per video dropping is real but the part most people skip is that cheap production only matters if the hook converts. A good tip: use view-based diagnostics before scaling volume. If a video sits at 0-200 views it's a hook problem, not a budget problem. Producing 550 ads that all have weak hooks is just burning $2,750 faster. Better to batch 10-15 hook variations first, find the 2-3 that break 2K+ organically, then scale those with mutations
English

Great breakdown. One thing worth adding on the diagnostics side - if a post gets 0-200 views, that's almost always a hook failure, not an account problem. 200-2K means the story isn't holding. 2K-10K means engagement is weak (no comments/shares driving distribution). Knowing which bucket you're in saves a ton of time because the fix is completely different for each. Also for slideshows specifically, 5-7 slides under 12 seconds total tends to be the sweet spot - anything longer and retention drops off a cliff 🔥
English

how to get millions of tiktok views in one week using openclaw
(the exact playbook):
what you need:
> openclaw (free, runs on any old laptop)
> openai api for image generation ($0.25-$0.50 per post)
> post scheduler (e.g. postiz) for tiktok posting
> 60 seconds of your time per post to add audio and publish
why tiktok slideshows:
tiktok's own data shows photo carousels get 2.9x more comments and 2.6x more shares than video.
the algorithm is pushing them hard. and they're easy for ai to generate.
one guy got 1.2 million tiktok views in one week promoting a calorie tracking app. zero manual content creation.
the 6-slide formula:
> slide 1: hook with text overlay
> slide 2: the problem
> slide 3: discovery
> slide 4: transformation 1
> slide 5: transformation 2
> slide 6: cta
say you're promoting a calorie tracking app. this would look like:
> slide 1: "my nutritionist looked at what i actually eat in a day and went silent"
> slide 2: a messy kitchen counter with takeout boxes and snacks
> slide 3: someone scanning food with their phone
> slide 4: a clean meal prep layout with calories visible
> slide 5: a weekly breakdown showing progress
> slide 6: "i didn't change what i eat. i just started seeing the numbers"
the IMPORTANT thing you need to add to prompts that makes it work:
add "iphone photo" and "realistic lighting" to every prompt.
the images need to look like someone took a real photo on their phone.
then lock the architecture. write one obsessively detailed description of the scene.
counter dimensions, lighting direction, camera angle, what's on the table, everything. keep that identical across all 6 prompts. only change what's on screen or the food layout.
without this every slide looks like a completely different scene and nobody believes it.
text overlay rules:
> font size: 6.5% of image height with TIKTOK SANS font (which is free)
> position: 30% from top (top 10% hidden by tiktok status bar)
> line breaks every 4-6 words so text doesn't get squashed
> full hook on slide 1. never split across slides
> bottom 20% is hidden behind tiktok ui
the hook that actually works:
every hook that flopped was feature-focused:
> "track your calories in seconds" = dead
> "the best macro tracking app of 2026" = dead
what hits:
> "my trainer asked me to log everything i eat for a week. she wasn't ready for xyz"
> "showed my girlfriend how many calories are in her 'healthy' smoothie. she hasn't spoken to me since"
> "my doctor told me to track my food for 30 days. here's what happened by day 12"
the formula:
[another person] + [doubt or conflict] + you showed them the data + their reaction
relatability is everything. always include another person.
make sure that your agent has a learning loop:
your agent doesn't just post and forget. it:
> tracks views on every post
> logs which hooks performed and which flopped
> updates its own rules automatically
> generates new hooks based on winners
> checks if views are actually converting to downloads
high views no downloads = wrong audience, pivot hooks
downloads no paid users = fix onboarding or paywall
low views high conversions = hook needs work but cta is perfect
what to expect:
> week 1: few hundred to few thousand views. agent is learning
> week 2-3: posts start breaking 10k-50k
> month 2+: consistent 50k-100k+ with compounding
important tips:
> warm up your tiktok for 2-3 days before posting. scroll your niche. like sparingly. follow accounts. don't skip this so that you dont get 0 views
> always add trending audio manually. 10x reach difference over random music. tiktok will give it to you automatically
> post 3 times a day minimum
> let the agent learn. don't micromanage every image
> content performs best when it looks messy and real. not perfect.
your daily time: 60 seconds to open drafts, pick a sound, publish.
the agent handles everything else.
English

200 accounts is legit scale but the bottleneck nobody talks about is warmup and device isolation. If accounts share any fingerprint overlap - same Apple ID history, same proxy subnet, even the same WiFi - TikTok links them and one ban cascades across the whole batch. The safer rule is more accounts with fewer posts per account rather than maxing out posts on fewer accounts. Keeps each one under the radar longer
English

The short clips approach is key here. Keeping each image under 2-3 seconds makes AI detection way harder - both for viewers and for the platform itself. Slideshows with 5-7 frames and under 12 seconds total tend to hit the sweet spot. One thing to watch though: if the first video pops but the next 5 don't, check the retention curve. Sharp drop at 2 seconds = hook problem, slow decay = pacing issue. Diagnose before you just keep posting more
English

AI actually makes it so much easier to go viral
this TikTok account got 321k views on his first video only uploaded 16 hours ago
this proves quality of the video over just pushing out quantity that’s shit
they used images and very short clips which is the key with AI creatives
you give them less time to process the images It’s much harder for people to detect it’s AI
they pushed the traffic to an ecom product but if you can get creative and use this for sweeps you can literally make $10k a month pretty easily
AI creatives + organic affiliate is the best way for beginners to print hands down

English

Solid setup. One thing worth adding - skip Apple ID altogether during initial device setup if possible, and create it fresh only after connecting the proxy. If TikTok sees any Apple ID history from previous owners it can silently flag the device. Also for warmup, a good rule of thumb: 5 days of just lurking and liking before posting anything, then check if your first video crosses 700 views organically. If it doesn't, the account trust is probably low and it's cheaper to start over than to grind it out
English

This is how I try to target a US audience on TikTok step by step 🇺🇸
1. Buy cheap used iPhones on eBay ($25–35) and do a full factory reset
2. Set up the iPhone in English (US), region United States (skip Apple ID, turn off location, tracking, Siri and Screen Time)
3. Open Safari and create a fresh email (Gmail or Outlook, ...)
4. Create a new Apple ID with that email
5. Set up a VPN of your choice (NordVPN, Hetzner VPS, mobile proxy – everyone recommends something different) and download it from the App Store
6. Turn on the VPN – then make sure in iPhone Settings: General → Language & Region → United States and General → Date & Time → disable "Set Automatically" → set to a US city
7. Download TikTok and sign up with the fresh email – quick check: tap "phone number" first, if it shows +1 TikTok sees you as a US user ✅ then switch back to email and finish signing up
8. Warm up account for 7 days

English

cartoon characters in absurd situations crush it because curiosity retention is insane on those. One thing to keep in mind though: this format has a short lifecycle. Once 50 accounts copy the same angle, TikTok starts suppressing it. The move is to stack 5-7 variations of the concept fast (different scenarios, different characters) and track which ones hold above 700+ avg views. Once you see 6 consecutive underperformers, that variation is dead - rotate to the next one
English

100% this. One tip though - these style ads have a fast lifecycle. They blow up, everyone copies, platform catches on. The move is to batch 5-10 visual variations per winning angle right away and rotate hard. Once you see 6+ consecutive drops in performance on a variation, kill it and swap in the next one. Speed of iteration matters more than polish here
English

200+ videos/day is the right volume but the bottleneck nobody mentions is account health. Pushing that many videos across accounts without proper warmup (5 days lurking + algo training before first post) means most of them land under 300 views and get silently throttled. The real diagnostic: if test slideshows don't hit 700+ views within 48 hours, the account is already compromised. Better to kill it early and rotate than keep feeding dead accounts 💀
English

$351K in 39 days.
No team. No agency.
Just 8 AI creators + 1 repeatable system.
🔁 Find a hot product
⚡ Launch a simple Shopify store
🎯 Use CAA
📈 Push 200+ UGC videos/day
One product. Multiple angles. Scale fast.
Comment “CAA” for the strategy.
Must following (@monicaa_AI)

English

200+ videos/month means roughly 7/day - that's where posting cadence gets tricky. Most platforms start throttling if you push too many from one account. The move is splitting across 5-6 accounts in the same niche and staggering post times across US waking hours. Horizontal scaling (more accounts, fewer posts each) always outperforms vertical (one account, max posts). Also never use scheduling tools - post natively from camera roll only, metadata from third-party apps is a flag 🫡
English

This AI-powered TikTok page made $1,623,177 in 93 days.
No filming. No cooking. No face.
Just:
- 200+ AI recipe videos/month
- Faceless voiceovers + AI food images
- $25 cookbooks sold on autopilot
- 55M+ views | Dozens of 1M+ videos
It looks like a real food creator — but it’s 100% AI.
Want the full system comment “Page” and “Retweet” & I’ll send it
(Must be following)

English

Consistent faces is the hardest part at scale. Most AI tools introduce micro-variations between generations - slightly different eye spacing, skin tone shifts, jawline drift. The audience doesn't consciously notice but engagement slowly drops because something feels "off." Biggest unlock is controlling the seed + using reference images locked to one exact face across all outputs. That's what separates a real-looking page from one that feels AI-generated after 3 posts.
English

Al is everywhere now.
But very few people are doing it right.
This is how I create Al UGC that actually feels real.
Consistent faces.
Natural light.
Smooth, believable motion.
No uncanny vibes. No overdone styling.
Al influencers aren't coming next.
They're already here.
If you want to learn, comment UGC Or click on the link pinned in the comments and I'll break it down.
English

🧵 TikTok Farm Playbook — Part 4/20: Proxies & network (the only 2 things TikTok checks and why fraud scores are useless)
⬆️ Previously: One Apple ID per device, never reuse anything, clean creation SOP with 8 steps
Your network setup is your farm's passport. Get it wrong and nothing else matters - not your content, not your warmup, not your hooks.
Proxy ranking (non-negotiable):
🥇 Mobile proxies - carrier-backed IPs, mimic real mobile users. Safest at scale
🥈 Residential ISPs - acceptable if ASN + GEO are clean, don't rotate aggressively
🥉 VPNs - NEVER. Shared datacenter ASNs. TikTok associates them with automation instantly
Now here's what most people get wrong:
They check IP fraud scores (Scamalytics, IPQualityScore, Fraudlogix) and think "low score = safe."
❌ Wrong. Zero correlation with TikTok reach.
We had a viral account running on an IP with fraud score 69. And dead accounts on "clean" IPs with score 5.
TikTok checks exactly TWO things:
1️⃣ ASN - who owns the connection (must resolve to target country)
2️⃣ GEO - where the IP geolocates (must resolve to target country)
ASN = your passport. GEO = your face. TikTok cross-checks both. Mismatch = suspicion.
How to verify:
- bgp.he.net → check ASN country
- geo.brdtest.com/mygeo.json → check GEO country
Both must match. Both must match your target. Check during: app install, account creation, daily usage, AND posting.
Targeting US from outside the US? You need:
✅ US ASN + US GEO
✅ US timezone on phone (Eastern or Pacific)
✅ English system language
✅ Scroll/post only during US waking hours
❌ NOT needed: US SIM, US phone number, physical US presence
Trusted providers: Bright Data (dedicated ISPs), Decodo/Smartproxy (mobile proxies).
⬇️ Next: The warmup system - the most critical step that 90% of people skip (and why your first 5 days determine everything)
🔁 Repost if useful. Follow for the full series.
English







