Explore posts by Alex Nguyen
3,178 posts by Alex Nguyen
Select a post to readHCMC and Da Nang offer different tradeoffsinsightAs a Vietnamese person, I think I can add a bit of context to this.
Your experience in Da Nang is actually something I’ve heard from some foreigners before.
HCMC and Da Nang are very different cities.
HCMC has a lot of international companies and expats, so people there are more used to foreigners and English is more common.
It’s also known for having the best service culture in Vietnam.
But the downside is obvious:
traffic, noise, and air quality are worse because it’s a fast-growing megacity.
Da Nang is almost the opposite.
Pros:
- beautiful city
- great beach and sunshine
- much better air quality
- cheaper cost of living
But the trade-offs are:
- smaller city
- fewer English speakers
- service culture not as strong as HCMC
- fewer expat-heavy neighborhoods
In HCMC you have areas like Thao Dien, District 1, District 7 where locals interact with foreigners every day, and many foreigners live and work there long-term.
Da Nang is much smaller, and foreigners are more spread out, so locals may feel more reserved, not necessarily unfriendly.
Ah, one more thing I personally love about HCMC:
It’s basically a city that never sleeps.
You can find 24/7 cafes where people are coding, building startups, studying, or working late at night.
There are also 24/7 food spots, so you can order something anytime.
That “always moving” energy is something I really enjoy about living in HCMC.
So I’d summarize it like this:
HCMC → chaotic but efficient, city that never sleeps
Da Nang → beautiful, slower, and more relaxed
Both are great cities, just very different vibes.
Maybe in the future service in Da Nang will improve, and if you come back you might have a much better experience. 😁
Alex NguyenMar 2615450Validate distribution before building the productinsightI like this way of thinking:
Don’t build something just because you’re passionate about it while users don’t actually need it.
And don’t blindly build random apps either.
Find the distribution angle first.
My workflow:
> Find viral content in a specific niche
> Study the hooks, formats, pain points, and comments
> Identify what users are already interested in
> Check whether the niche can monetize through subscriptions or one-time purchases
> Test different content angles before building
> Build the simplest app that solves the problem
> Create content using the formats that already proved they can go viral
> Track views, installs, conversion, retention, and revenue
> Improve the product based on real user behavior
> Scale the best-performing channels with more organic content or paid ads
Distribution first. Product second.
Build the product around a proven path to users.
It reduces risk and significantly increases your chances of success.Alex NguyenJun 26384683Pinterest Claude ArcadsguidePinterest + Claude + Arcads
That's all you need to build an AI influencer farm to scale your apps.
Let me show you how I did it:
1. Go to Pinterest and grab a photo you like (pose, vibe, lighting)
2. Upload it to Claude and ask:
“Give me a detailed JSON prompt for this image holding [your product] or reviewing my B2C app or anything”
3. Open NanoBanana Pro and paste the prompt → upload your product image (white background works best)
4. Take the generated image and animate it with Kling 3.0 inside Arcads
Done.
You’ve just created an AI influencer farm in <5 minutes.
No creators. No studio.
And no $1k per UGC video.
Organic > ads.
Scale content, not headcount.
PS: Huge thanks to Arcads for sponsoring our efforts 💪Alex NguyenFeb 26333557The simplest way to create an AI influencerinsightThe simplest way to create an AI influencer using workflow (3 steps):
Nano Banana → Veo 3 → Arcads AI Workflow
Step 1: Build your character with a JSON prompt (Nano Banana Workflow)
→ Use a structured JSON prompt instead of writing random text
→ Define the look clearly: character, outfit, lighting, camera style, background
→ Add negative constraints like: no logos, no text, no extra fingers, no CGI skin, no warped mirror
This makes the output more stable and reusable.
You’re not just prompting.
You’re building a repeatable image system
Step 2: Turn image → video (Veo 3 Workflow)
→ Connect the Nano Banana output directly into Veo 3
→ Add a simple motion prompt: slight body movement, natural head turn, subtle phone movement
→ Keep it minimal so it feels more real
Workflow removes the manual back-and-forth.
Step 3: Add script + voice (Arcads AI Workflow)
→ Feed the generated video into Arcads AI
→ Add a short script, usually 10–20 seconds
→ Generate voice + lipsync that matches the character vibe
Now the character becomes actual content.
Why workflow matters:
Prompt → Image → Video → Voice
No messy manual process.
No repeating the same setup every time.
Just a clean pipeline you can scale.
Nano Banana + Veo 3 + Arcads AI + workflow feature
= simple AI influencer factoryAlex NguyenMar 26227468AI built Star Defender in five hoursinsightAfter I posted 2 videos about the game I built entirely with AI, a lot of people didn’t believe it was real.
Well, here’s the video showing how I built the full Star Defender game using Claude Fable 5 + Rork.
In total, it took me around 5 hours: about 1 hour to build the gameplay, and another 4 hours to turn it into an almost complete game. It also cost around 1,500 Rork credits, which is about $300 based on Rork’s pricing.
Also, Rork already has built-in integrations for generating game art, sound effects, code..., so I didn’t use any external tools at all.
As I mentioned in my previous video, I wanted to test Claude Fable 5 to make a game, and since I’m a Rork partner, I had free credits to use.
I also only used natural language prompts, mixing both Vietnamese and English. There were no complex prompts or prompt engineering involved.
Yes, I only prompted. I didn’t touch any code. I was basically just acting as a game tester: playing the game, taking screenshots whenever I found bugs, and describing them so the AI could fix them.
And honestly, it performed better than I expected.
At first, I only wanted to test how well Fable 5 could build games. But now, I want to experiment with building a game from scratch all the way to releasing it on the App Store, including AdMob integration and in-game purchases to monetize it.Alex NguyenJul 26152121AI influencers could reshape organic marketinginsightTikTok / IG / YT AI Influencers will be one of the biggest trends in 2026.
I’m very confident about this.
Think about it 👇
Instead of paying $1k–$10k for ONE video from a “big” influencer
→ with zero guarantee it goes viral
Instead of:
- booking multiple influencers
- burning money on high CPMs
- praying one post hits
You can now:
→ Create multiple AI influencers
→ Automate ~80% of the content pipeline
→ Control hooks, angles, frequency, distribution
Why not 100% automated?
Because:
- posting still needs to look human
- accounts must be warmed up
- full automation = spam + shadowban
Platforms don’t hate AI.
They hate low-effort automation.
If you crack:
- TikTok AI influencers
- IG / FB Reels / YouTube Shorts AI influencers
You’re not “early.”
You’re already operating in 2026 marketing.
For small businesses with limited budget:
Organic scale > paid influence.Alex NguyenJan 2611281Grok built a complex strategy gameguideI just tested Grok 4.5 in Rork to build a strategy city-building and survival mobile game with complex logic.
I only described the core idea.
It generated the full game:
> Art
> Gameplay
> UI/UX
> Sound
> Game logic
> Controls
> In-game economy
> Balancing
The result was much better than I expected.
Honestly surprised Grok 4.5 handles game logic and gameplay controls on the same level as Fable 5 and GPT-5.6 Sol.
AI game development is moving insanely fast. 🥴Alex NguyenJul 266820A fast path to $10K MRRguideHow to hit $10K MRR in 2026 (fast):
- Build a B2C app that solves a painful problem
- Pick a Gen Z market: ADHD, sleep, mental health, fitness...
- Use real people, real stories, real trends, authenticity wins
- Post every day on TikTok, IG, Threads, and X. No zero days
- Every post needs: a strong hook + real value + clear CTA
- Build an AI Influencer Factory to multiply organic distribution
- Once organic works, pour fuel on it with paid ads
Do this for 30 days.
You don’t need 100 viral posts.
You need one.
One viral hit → traffic → installs → revenue.
In 2026, distribution is the moat.
Ship fast.
Market faster.
Repeat.
PS: Actually, if you want the fastest way, you can just drop $20k on paid ads lol.Alex NguyenSep 26684973Create AI influencer videos with NanoBanana and ArcadsguideJust cracked a stupid-simple way to create AI influencer content 👀
NanoBanana + Arcads
That’s literally my entire stack.
How it works:
1. Find a trending video on TikTok / IG / YT
2. Generate a fresh AI influencer with NanoBanana
3. Use Arcads to turn it into an AI influencer video
You’re not copying.
You’re remixing what already works.
That’s smart distribution.
This is exactly why AI influencer farms are exploding in 2026.
Organic scale > paid influencers when you’re small.
PS: Huge thanks to Arcads for sponsoring our experiments 💪
The prompt is below 👇Alex NguyenJan 26461960A solo founder path to a million-dollar exitguideNew 2026 playbook for solo founders & indie hackers:
If you want to build a $1M+ exit (or even a multi-million dollar portfolio), stop chasing random ideas.
1. Go to SensorTower.
- Find apps doing $1M+/month.
(Real revenue is usually 1.5–2x higher.)
2. Vibe coding it or build it yourself, it doesn’t matter. Just make sure it’s a better.
- Doesn’t need to be perfect.
- Just needs to be better.
3. Distribute on TikTok, IG, YouTube, X.
- Attention > product.
4. Organic first.
- Then scale with influencers + ads.
- Use OpenClaw to automate everything you can.
5. Get acquired by the big players you studied.
Retire.
Or repeat the cycle.
PS: I’m on that path too.Alex NguyenMar 26556957Stack distribution channels to grow an appguideIf you want to grow an app to $10k/month, distribution matters more than the product.
From my experience, the 4 best beginner-friendly marketing channels are:
1. Organic TikTok slideshows
- Easy to make. Can get millions of views.
- But not always the best for paid conversion.
2. Organic AI influencers/UGC
- Better conversion, but expensive and hard to scale.
- You need a real workflow before building an AI UGC army.
3. Hiring paid influencers/UGC
- Can work really well.
- But creator quality is inconsistent, outreach takes forever, and good influencers are not cheap.
4. Paid ads (Meta Ads, Tiktok Ads, Google Ads, Apple Search Ads)
- The fastest way to scale MRR.
- But margins drop fast if your niche or funnel doesn’t convert well.
No channel is perfect.
The real unlock is stacking channels:
organic + influencers + AI UGC + paid ads.
That’s how you scale. 😌Alex NguyenJul 26629820Improve boring apps with better distributioninsightOne of the easiest plays right now if you want to build a $10k/month app but feel like every idea already exists:
Don’t invent. Improve or repackage.
→ Find a boring app making $1M+ (Sensor Tower is usually inaccurate for small apps, but much more reliable for big ones because there’s more data)
→ Clone it, Vibe coding no matter
→ Make it cleaner, funnier, gamified ( Duolingo for ... )
→ Push organic first on social media Tiktok, IG, Threads, X...
→ Then scale with paid ads
Boring apps print money.
Creative distribution makes them explode 😌Alex NguyenApr 26351516Four tradeoffs in app distribution channelsguideMust-read if you want to promote an app or sell any product on TikTok or IG.
And here are a few updates based on my experience. The 4 best methods I’ve seen for beginners are:
1. TikTok slideshows
Easy to make and capable of getting high views, but harder to convert into paying users.
2. AI influencers
Usually convert better, but they’re quite expensive and take a lot of time to build a proper workflow and scale an AI UGC army.
3. Hiring influencers
Works pretty well, but the downside is that it costs a lot of money. Sometimes influencers put in minimal effort, so you need strong quality control. It’s also very time-consuming to DM people individually, search for, and filter high-quality influencers.
4. Paid ads
The biggest advantage is that you can scale MRR very quickly. But the downside is lower profit margins, and you need to choose the right app niche with high conversion rates to stay profitable while running ads.
Each method has tradeoffs.
The best strategy is usually combining multiple distribution channels together. 😌Alex NguyenMay 26257488This is my TikTok growth playbook every B2CguideThis is my TikTok growth playbook every B2C founder should know
What actually works (with me) in 2024-2026:
> TikTok first.
The algo is totally different from IG or YT Shorts.
Small accounts can still go viral → distribution is real.
> Quantity > perfection.
Work with 25–50 micro creators (<10k subs).
Post 1–2 videos daily. Test fast.
> Cross-post everything.
TikTok → IG Reels → FB → YT Shorts.
Same content, more surface area.
> Don’t reinvent formats.
Clone what already goes viral.
Strong 3-second hook or you’re dead.
> Find 1 viral format → double down.
Post it repeatedly until it dies, then move on.
> Hook + CTA must align.
Otherwise you’ll get millions of views
and 0 downloads.
> Pay for views, not upfront.
Align incentives.
Keep only creators that convert.
> Track everything.
UTMs → views → clicks → installs.
Cut fast. Scale faster.
> Warm new accounts.
Seed them with non-marketing content first
before pushing ads or product videos.
> Reality check:
Not every product works on TikTok.
Best way to find ideas?
Search viral pain points on TikTok / IG / YT →
then build an app that solves that exact pain.
Marketing > coding.
Test fast. Learn faster. Ship what converts. 🫡Alex NguyenFeb 26214405A low-capital path to app revenueguideIf you’re an indie hacker starting with little (or zero) capital, like I did:
- Don’t reinvent the wheel.
- Find an app with a proven market
- Build a cleaner, more polished version
- Add one or two clear differentiators
- Market it where competitors are lazy or absent
- This gets you to $5k–$10k MRR way faster.
Once cash flow is stable → then go deep on differentiation and scale.
Simple. Boring. It works.Alex NguyenFeb 26468363A lot of people ask me how I’mguideA lot of people ask me how I’m able to create slideshow content that gets millions of views, produce a large volume every day, and still keep the quality high:
> Step 1: Scroll TikTok within the niche your app is focused on.
For example, I focus on “studytok” for education apps.
> Step 2: Pick a piece of content that’s already viral and put it into ChatGPT to recreate a new version while keeping the same core idea and hook.
Don’t reuse the exact hook from the viral post, TikTok can easily detect it as copied content, which may lead to shadow banning.
> Step 3: Find relevant images and use Canva to design the slideshow.
> Step 4: Post at the time when users are most active in the country you’re targeting.
> Step 5: If the content doesn’t work, repeat the steps above.
This is how I used to do everything manually.
Now, I use my own internal AI content factory to run this process at scale, both for myself and for the students I’ve hired and manage.Alex NguyenJan 26171350VPN setup matters more than VPN locationguideToo many people keep asking me:
“Does using a VPN get you shadowbanned on TikTok?”
Short answer: No.
TikTok does not ban VPNs.
But most people use them wrong.
A few things you must know:
> Don’t use cheap or low-quality VPNs shared by too many users
→ Imagine TikTok seeing 1,000 accounts on the same IP
> Always turn on your VPN before opening TikTok
→ You don’t want your IP jumping between continents
> To target US users, you don’t have to pick a US VPN
→ Countries in the same time zone like Canada work just fine
> The US has massive communities speaking Spanish and other languages
→ Try creating content in those languages
→ Less competition, easier reach
VPN ≠ shadowban
Bad setup = problemsAlex NguyenJan 26236318Avoiding zero views with a TikTok setup seriesguideI’m thinking about remaking a free step-by-step series on how to set up accounts and find content to avoid getting 0 views.
I’ve noticed that many people are getting very low views, whether they use a VPN or not.
Meanwhile, on the 3 phones below, I only started posting less than 2 months ago. There are 6 accounts in total, and none of them are shadow banned. The average is around 10K views per post, with quite a few posts getting over 100K views. All phones are using VPNs.
This has helped me get an average of 700 new users per app every day.
So I think if your views are flopping, it’s likely because you set up something wrong, your content has visuals that violate TikTok’s policies, your IP is blacklisted, or simply your content isn’t good enough for TikTok to keep pushing it viral.Alex NguyenJun 26318304A workflow for multiple AI influencer variantsguideHow to create multiple AI Influencer variants (step-by-step) using Claude Opus 4.7 + Arcads Workflow.
This is how you scale content WITHOUT hiring creators 👇
Step 1: Find a viral reference
Go to TikTok
→ search your niche (fitness, study, looksmaxing, etc.)
→ find 1 influencer video with high views
You don’t need originality
You need what already works
Step 2: Turn image → JSON → variants
Take screenshots from that video
Drop into Claude Opus 4.7
Prompt it like:
“Convert this person into a structured JSON prompt for image generation + create 5–10 variants (different face, hair, pose, vibe)”
Now you have:
1 base character
5–10 variations
→ same identity, different “angles”
This is your UGC army
Step 3: Scale images with Arcads
Inside Arcads:
Create a workflow
Paste each JSON variant from Claude
Add a nanobanana image block (or any image gen model)
Duplicate block = number of variants
Now you’re generating:
→ multiple influencers
→ in parallel
→ fully automated
Step 4: Turn images → video
Create ONE master prompt:
“Generate a short-form video of this character posing / walking / talking with subtle motion, cinematic lighting, viral TikTok style”
Then:
Connect all image outputs
→ into 1 AI video generation block
Now you get:
→ multiple videos
→ same format
→ different characters
Why this works
You’re not creating content
You’re building a system that prints content
Variants = more hooks
More hooks = more chances to go viral
More viral = more installs / revenue
Real insight most people miss:
1 influencer = 1 chance
10 variants = 10 chances
Distribution beats perfection.Alex NguyenApr 26125301A repeatable playbook for B2C appsguideB2C App Playbook (2026):
1. Ship fast → weeks, not months. Ugly > perfect.
2. Pick ONE channel → TikTok, X, or Reddit.
3. Consistency > luck → 100 posts beats 1 viral spike.
4. Charge early → $5/month > 5,000 free users.
5. Retention compounds → growth hacks don’t.
6. iOS + Web first → that’s where the money is.
7. Win organic first → paid comes later.
Simple.
Brutal.
Repeatable.
If you want it edgier or shorter for max reach, say the word and I’ll tune it.Alex NguyenJan 26279258Improve proven products instead of copying blindlyinsightMust-read if you’re trying to build a successful app.
I really like this part: “Stop avoiding copying what works from other people. Doing this is actually wise product thinking.”
Because if all you do is copy another app step by step the product, features, and even their marketing strategy you’re basically building for them and helping market their product.
Over time, your product thinking will die. You lose creativity, become dependent on what others are doing, and will always be one step behind.
Instead, learn from what they’re doing right, understand why it works, and apply those lessons to your own product in your own way.
Learn from what works. Don’t blindly copy it.Alex NguyenAug 26122249TikTok uses local signals to target viewersguideLesson 2: You need a VPN because TikTok’s algorithm first targets local viewers. Try checking your audience traffic in TikTok Studio to see it yourself.
If you want to target users in the US while you’re in Vietnam like me, a VPN or proxy is necessary. If you’re already in the US, you don’t need one unless you want to scale with 10–30–100 accounts, in which case you’ll need it to avoid spam bans.
As for automation, if your account is already trustworthy and has lot of followers, it can work. But if it’s a completely new account and you use automation tools right away, I’m not so sure.Alex NguyenMar 26179249A short onboarding hurt app conversioninsightMy first failure of 2026.
I was very confident about my AI calories app, but I made a mistake.
The onboarding was too short and the core features didn’t stand out enough compared to other calorie tracking apps.
I started organic TikTok marketing and it actually worked pretty well. Traffic was coming in.
But almost 100% of users started the trial and then canceled. Conversion = 0.
So I decided to stop marketing for now and focus on shipping v2.0.0 to make the product much better.
Sometimes the problem isn’t distribution.
It’s the product. 🥴Alex NguyenMar 26509224Daily posting is the simplest marketing strategyguideIf you don’t know how to market, do this:
> Post every day
> Use a clean, trustworthy profile photo
> Make your bio stupidly clear (what you do + who it’s for)
> Reply to 100+ people (real humans, not AI spam)
> Connect with new creators daily
You’ll learn 10× faster by doing than by overthinking.
Show up every day.
That’s the simplest marketing strategy on earth.Alex NguyenJan 26365207Human-like TikTok workflows can still go viralinsightAnother guy followed the free sauce I shared and went viral with 300K+ views after posting consistently for 45 days and 98% US viewers (higher than mine lol 😭). Congrats, bro.
Shadow-bans are a joke. If you’ve only posted for a few days and haven’t gotten high views, change your content format before abandoning the account.
As I mentioned in my previous post, TikTok increases the risk of bans when you automate and your posting behavior looks unnatural. But if your workflow behaves like a real human user, TikTok usually won’t ban you, even if you use a VPN or proxy.
That said, a bad IP that has been heavily spammed on TikTok is still not good.Alex NguyenJul 26124189Most B2C founders overthink marketing.guideMost B2C founders overthink marketing.
This is the path I’ve seen work again and again:
$0–$20k
→ Post organic content yourself
→ Build a small AI content factory
$20k–$100k
→ Hire UGC creators to scale
→ Still keep posting organic personally
$100k+
→ Turn on paid ads
→ Scale without artificial limits
Don’t rush ads.
Organic creates leverage.
Paid ads just amplify what’s already working.Alex NguyenFeb 26206180Production basics separate apps from demosguideEvery vibe coder should understand this (Part 2) 👇
Stuff that separates “it runs on my laptop” from “real product”:
Shipping & prod basics:
- dev vs staging vs prod ≠ same thing
- build passes locally ≠ works in prod
- logs > guessing (read them)
- crashes in prod are normal, silence is not
- rollbacks > panic fixes
- config changes can break everything
APIs & integrations:
- rate limits exist (you WILL hit them)
- API keys ≠ frontend code
- retries + timeouts matter
- 3rd-party services WILL go down
- mock data before real APIs
Frontend survival:
- loading states aren’t optional
- empty states are UX, not bugs
- errors should be user-friendly
- mobile first (yes, really)
- slow UI = users bounce
Founder mindset shift:
- perfection blocks shipping
- boring solutions make money
- speed > elegance (early on)
- users don’t care about your stack
- shipping > tweeting about shipping
You’re not learning to code.
You’re learning to finish things.Alex NguyenFeb 26153160Find a winning format before automating TikTokguideEveryone wants to automate TikTok.
But automation should come AFTER you find a winning format.
Especially if you’re targeting the US with VPN/proxy.
My rule:
Test manually → find repeatable views → break down why it works → turn it into a template → automate the boring parts.
Automation won’t fix a bad format.
It only scales what already works.Alex NguyenJun 26148146Vibe coders need basic production safeguardsguideEvery vibe coder needs to understand this 👇
Not advanced engineering.
Just the boring basics that save you from breaking everything.
Non-negotiables:
> .env = secrets live here
> Never push .env to GitHub
>npm run dev = local only
> package.json = scripts, dependencies, versions
> npm run build = what actually ships
> npm packages = install / remove / update carefully
> git add .env = instant regret
> localhost:3000 ≠ a real URL
> env vars ≠ hardcoded values
> terminal errors are not decoration
Nice-to-know stuff that saves days:
> npm i --legacy-peer-deps when packages fight
> .gitignore = files Git should ignore
> basic Git: clone, pull, commit, push
> DB schemas / SQL if your app stores data
> auth basics: users, sessions, tokens
> dev words: components, modals, cards, routes
> reading docs > randomly guessing
> asking AI clear questions > “fix this bro”
You don’t need to be a “real dev.”
You just need to understand enough to ship without nuking your own app.
Happy vibe coding ✌️Alex NguyenMay 2697126Diagnosing TikTok posts stuck at zero viewsguideIf your TikTok gets 0 views, it usually means the video/slideshow hasn’t been pushed to the For You feed yet, or your account/video is being restricted.
For a new account, 0 views for the first few minutes or hours isn’t a big deal.
But if it’s still at exactly 0 after 12–24 hours, something is probably wrong:
→ Video is still under review/processing
→ Content looks like spam, a reupload, or duplicate content
→ New account posting too fast / behaving like a bot
→ VPN, proxy, or IP changes too frequently
→ Too many accounts on the same device with similar behavior
→ Video is accidentally set to Only Me/Friends
→ Account is temporarily restricted
If only ONE of your accounts gets 0 views → probably an account/video issue.
If MULTIPLE accounts suddenly get 0 views → check your device, IP, and posting workflow first.
Don’t keep spamming posts. It usually makes things worse.Alex NguyenSep 2610393
Alex NguyenMar 2615450Validate distribution before building the productinsightI like this way of thinking:
Don’t build something just because you’re passionate about it while users don’t actually need it.
And don’t blindly build random apps either.
Find the distribution angle first.
My workflow:
> Find viral content in a specific niche
> Study the hooks, formats, pain points, and comments
> Identify what users are already interested in
> Check whether the niche can monetize through subscriptions or one-time purchases
> Test different content angles before building
> Build the simplest app that solves the problem
> Create content using the formats that already proved they can go viral
> Track views, installs, conversion, retention, and revenue
> Improve the product based on real user behavior
> Scale the best-performing channels with more organic content or paid ads
Distribution first. Product second.
Build the product around a proven path to users.
It reduces risk and significantly increases your chances of success.Alex NguyenJun 26384683Pinterest Claude ArcadsguidePinterest + Claude + Arcads
That's all you need to build an AI influencer farm to scale your apps.
Let me show you how I did it:
1. Go to Pinterest and grab a photo you like (pose, vibe, lighting)
2. Upload it to Claude and ask:
“Give me a detailed JSON prompt for this image holding [your product] or reviewing my B2C app or anything”
3. Open NanoBanana Pro and paste the prompt → upload your product image (white background works best)
4. Take the generated image and animate it with Kling 3.0 inside Arcads
Done.
You’ve just created an AI influencer farm in <5 minutes.
No creators. No studio.
And no $1k per UGC video.
Organic > ads.
Scale content, not headcount.
PS: Huge thanks to Arcads for sponsoring our efforts 💪Alex NguyenFeb 26333557The simplest way to create an AI influencerinsightThe simplest way to create an AI influencer using workflow (3 steps):
Nano Banana → Veo 3 → Arcads AI Workflow
Step 1: Build your character with a JSON prompt (Nano Banana Workflow)
→ Use a structured JSON prompt instead of writing random text
→ Define the look clearly: character, outfit, lighting, camera style, background
→ Add negative constraints like: no logos, no text, no extra fingers, no CGI skin, no warped mirror
This makes the output more stable and reusable.
You’re not just prompting.
You’re building a repeatable image system
Step 2: Turn image → video (Veo 3 Workflow)
→ Connect the Nano Banana output directly into Veo 3
→ Add a simple motion prompt: slight body movement, natural head turn, subtle phone movement
→ Keep it minimal so it feels more real
Workflow removes the manual back-and-forth.
Step 3: Add script + voice (Arcads AI Workflow)
→ Feed the generated video into Arcads AI
→ Add a short script, usually 10–20 seconds
→ Generate voice + lipsync that matches the character vibe
Now the character becomes actual content.
Why workflow matters:
Prompt → Image → Video → Voice
No messy manual process.
No repeating the same setup every time.
Just a clean pipeline you can scale.
Nano Banana + Veo 3 + Arcads AI + workflow feature
= simple AI influencer factoryAlex NguyenMar 26227468AI built Star Defender in five hoursinsightAfter I posted 2 videos about the game I built entirely with AI, a lot of people didn’t believe it was real.
Well, here’s the video showing how I built the full Star Defender game using Claude Fable 5 + Rork.
In total, it took me around 5 hours: about 1 hour to build the gameplay, and another 4 hours to turn it into an almost complete game. It also cost around 1,500 Rork credits, which is about $300 based on Rork’s pricing.
Also, Rork already has built-in integrations for generating game art, sound effects, code..., so I didn’t use any external tools at all.
As I mentioned in my previous video, I wanted to test Claude Fable 5 to make a game, and since I’m a Rork partner, I had free credits to use.
I also only used natural language prompts, mixing both Vietnamese and English. There were no complex prompts or prompt engineering involved.
Yes, I only prompted. I didn’t touch any code. I was basically just acting as a game tester: playing the game, taking screenshots whenever I found bugs, and describing them so the AI could fix them.
And honestly, it performed better than I expected.
At first, I only wanted to test how well Fable 5 could build games. But now, I want to experiment with building a game from scratch all the way to releasing it on the App Store, including AdMob integration and in-game purchases to monetize it.Alex NguyenJul 26152121AI influencers could reshape organic marketinginsightTikTok / IG / YT AI Influencers will be one of the biggest trends in 2026.
I’m very confident about this.
Think about it 👇
Instead of paying $1k–$10k for ONE video from a “big” influencer
→ with zero guarantee it goes viral
Instead of:
- booking multiple influencers
- burning money on high CPMs
- praying one post hits
You can now:
→ Create multiple AI influencers
→ Automate ~80% of the content pipeline
→ Control hooks, angles, frequency, distribution
Why not 100% automated?
Because:
- posting still needs to look human
- accounts must be warmed up
- full automation = spam + shadowban
Platforms don’t hate AI.
They hate low-effort automation.
If you crack:
- TikTok AI influencers
- IG / FB Reels / YouTube Shorts AI influencers
You’re not “early.”
You’re already operating in 2026 marketing.
For small businesses with limited budget:
Organic scale > paid influence.Alex NguyenJan 2611281Grok built a complex strategy gameguideI just tested Grok 4.5 in Rork to build a strategy city-building and survival mobile game with complex logic.
I only described the core idea.
It generated the full game:
> Art
> Gameplay
> UI/UX
> Sound
> Game logic
> Controls
> In-game economy
> Balancing
The result was much better than I expected.
Honestly surprised Grok 4.5 handles game logic and gameplay controls on the same level as Fable 5 and GPT-5.6 Sol.
AI game development is moving insanely fast. 🥴Alex NguyenJul 266820A fast path to $10K MRRguideHow to hit $10K MRR in 2026 (fast):
- Build a B2C app that solves a painful problem
- Pick a Gen Z market: ADHD, sleep, mental health, fitness...
- Use real people, real stories, real trends, authenticity wins
- Post every day on TikTok, IG, Threads, and X. No zero days
- Every post needs: a strong hook + real value + clear CTA
- Build an AI Influencer Factory to multiply organic distribution
- Once organic works, pour fuel on it with paid ads
Do this for 30 days.
You don’t need 100 viral posts.
You need one.
One viral hit → traffic → installs → revenue.
In 2026, distribution is the moat.
Ship fast.
Market faster.
Repeat.
PS: Actually, if you want the fastest way, you can just drop $20k on paid ads lol.Alex NguyenSep 26684973Create AI influencer videos with NanoBanana and ArcadsguideJust cracked a stupid-simple way to create AI influencer content 👀
NanoBanana + Arcads
That’s literally my entire stack.
How it works:
1. Find a trending video on TikTok / IG / YT
2. Generate a fresh AI influencer with NanoBanana
3. Use Arcads to turn it into an AI influencer video
You’re not copying.
You’re remixing what already works.
That’s smart distribution.
This is exactly why AI influencer farms are exploding in 2026.
Organic scale > paid influencers when you’re small.
PS: Huge thanks to Arcads for sponsoring our experiments 💪
The prompt is below 👇Alex NguyenJan 26461960A solo founder path to a million-dollar exitguideNew 2026 playbook for solo founders & indie hackers:
If you want to build a $1M+ exit (or even a multi-million dollar portfolio), stop chasing random ideas.
1. Go to SensorTower.
- Find apps doing $1M+/month.
(Real revenue is usually 1.5–2x higher.)
2. Vibe coding it or build it yourself, it doesn’t matter. Just make sure it’s a better.
- Doesn’t need to be perfect.
- Just needs to be better.
3. Distribute on TikTok, IG, YouTube, X.
- Attention > product.
4. Organic first.
- Then scale with influencers + ads.
- Use OpenClaw to automate everything you can.
5. Get acquired by the big players you studied.
Retire.
Or repeat the cycle.
PS: I’m on that path too.Alex NguyenMar 26556957Stack distribution channels to grow an appguideIf you want to grow an app to $10k/month, distribution matters more than the product.
From my experience, the 4 best beginner-friendly marketing channels are:
1. Organic TikTok slideshows
- Easy to make. Can get millions of views.
- But not always the best for paid conversion.
2. Organic AI influencers/UGC
- Better conversion, but expensive and hard to scale.
- You need a real workflow before building an AI UGC army.
3. Hiring paid influencers/UGC
- Can work really well.
- But creator quality is inconsistent, outreach takes forever, and good influencers are not cheap.
4. Paid ads (Meta Ads, Tiktok Ads, Google Ads, Apple Search Ads)
- The fastest way to scale MRR.
- But margins drop fast if your niche or funnel doesn’t convert well.
No channel is perfect.
The real unlock is stacking channels:
organic + influencers + AI UGC + paid ads.
That’s how you scale. 😌Alex NguyenJul 26629820Improve boring apps with better distributioninsightOne of the easiest plays right now if you want to build a $10k/month app but feel like every idea already exists:
Don’t invent. Improve or repackage.
→ Find a boring app making $1M+ (Sensor Tower is usually inaccurate for small apps, but much more reliable for big ones because there’s more data)
→ Clone it, Vibe coding no matter
→ Make it cleaner, funnier, gamified ( Duolingo for ... )
→ Push organic first on social media Tiktok, IG, Threads, X...
→ Then scale with paid ads
Boring apps print money.
Creative distribution makes them explode 😌Alex NguyenApr 26351516Four tradeoffs in app distribution channelsguideMust-read if you want to promote an app or sell any product on TikTok or IG.
And here are a few updates based on my experience. The 4 best methods I’ve seen for beginners are:
1. TikTok slideshows
Easy to make and capable of getting high views, but harder to convert into paying users.
2. AI influencers
Usually convert better, but they’re quite expensive and take a lot of time to build a proper workflow and scale an AI UGC army.
3. Hiring influencers
Works pretty well, but the downside is that it costs a lot of money. Sometimes influencers put in minimal effort, so you need strong quality control. It’s also very time-consuming to DM people individually, search for, and filter high-quality influencers.
4. Paid ads
The biggest advantage is that you can scale MRR very quickly. But the downside is lower profit margins, and you need to choose the right app niche with high conversion rates to stay profitable while running ads.
Each method has tradeoffs.
The best strategy is usually combining multiple distribution channels together. 😌Alex NguyenMay 26257488This is my TikTok growth playbook every B2CguideThis is my TikTok growth playbook every B2C founder should know
What actually works (with me) in 2024-2026:
> TikTok first.
The algo is totally different from IG or YT Shorts.
Small accounts can still go viral → distribution is real.
> Quantity > perfection.
Work with 25–50 micro creators (<10k subs).
Post 1–2 videos daily. Test fast.
> Cross-post everything.
TikTok → IG Reels → FB → YT Shorts.
Same content, more surface area.
> Don’t reinvent formats.
Clone what already goes viral.
Strong 3-second hook or you’re dead.
> Find 1 viral format → double down.
Post it repeatedly until it dies, then move on.
> Hook + CTA must align.
Otherwise you’ll get millions of views
and 0 downloads.
> Pay for views, not upfront.
Align incentives.
Keep only creators that convert.
> Track everything.
UTMs → views → clicks → installs.
Cut fast. Scale faster.
> Warm new accounts.
Seed them with non-marketing content first
before pushing ads or product videos.
> Reality check:
Not every product works on TikTok.
Best way to find ideas?
Search viral pain points on TikTok / IG / YT →
then build an app that solves that exact pain.
Marketing > coding.
Test fast. Learn faster. Ship what converts. 🫡Alex NguyenFeb 26214405A low-capital path to app revenueguideIf you’re an indie hacker starting with little (or zero) capital, like I did:
- Don’t reinvent the wheel.
- Find an app with a proven market
- Build a cleaner, more polished version
- Add one or two clear differentiators
- Market it where competitors are lazy or absent
- This gets you to $5k–$10k MRR way faster.
Once cash flow is stable → then go deep on differentiation and scale.
Simple. Boring. It works.Alex NguyenFeb 26468363A lot of people ask me how I’mguideA lot of people ask me how I’m able to create slideshow content that gets millions of views, produce a large volume every day, and still keep the quality high:
> Step 1: Scroll TikTok within the niche your app is focused on.
For example, I focus on “studytok” for education apps.
> Step 2: Pick a piece of content that’s already viral and put it into ChatGPT to recreate a new version while keeping the same core idea and hook.
Don’t reuse the exact hook from the viral post, TikTok can easily detect it as copied content, which may lead to shadow banning.
> Step 3: Find relevant images and use Canva to design the slideshow.
> Step 4: Post at the time when users are most active in the country you’re targeting.
> Step 5: If the content doesn’t work, repeat the steps above.
This is how I used to do everything manually.
Now, I use my own internal AI content factory to run this process at scale, both for myself and for the students I’ve hired and manage.Alex NguyenJan 26171350VPN setup matters more than VPN locationguideToo many people keep asking me:
“Does using a VPN get you shadowbanned on TikTok?”
Short answer: No.
TikTok does not ban VPNs.
But most people use them wrong.
A few things you must know:
> Don’t use cheap or low-quality VPNs shared by too many users
→ Imagine TikTok seeing 1,000 accounts on the same IP
> Always turn on your VPN before opening TikTok
→ You don’t want your IP jumping between continents
> To target US users, you don’t have to pick a US VPN
→ Countries in the same time zone like Canada work just fine
> The US has massive communities speaking Spanish and other languages
→ Try creating content in those languages
→ Less competition, easier reach
VPN ≠ shadowban
Bad setup = problemsAlex NguyenJan 26236318Avoiding zero views with a TikTok setup seriesguideI’m thinking about remaking a free step-by-step series on how to set up accounts and find content to avoid getting 0 views.
I’ve noticed that many people are getting very low views, whether they use a VPN or not.
Meanwhile, on the 3 phones below, I only started posting less than 2 months ago. There are 6 accounts in total, and none of them are shadow banned. The average is around 10K views per post, with quite a few posts getting over 100K views. All phones are using VPNs.
This has helped me get an average of 700 new users per app every day.
So I think if your views are flopping, it’s likely because you set up something wrong, your content has visuals that violate TikTok’s policies, your IP is blacklisted, or simply your content isn’t good enough for TikTok to keep pushing it viral.Alex NguyenJun 26318304A workflow for multiple AI influencer variantsguideHow to create multiple AI Influencer variants (step-by-step) using Claude Opus 4.7 + Arcads Workflow.
This is how you scale content WITHOUT hiring creators 👇
Step 1: Find a viral reference
Go to TikTok
→ search your niche (fitness, study, looksmaxing, etc.)
→ find 1 influencer video with high views
You don’t need originality
You need what already works
Step 2: Turn image → JSON → variants
Take screenshots from that video
Drop into Claude Opus 4.7
Prompt it like:
“Convert this person into a structured JSON prompt for image generation + create 5–10 variants (different face, hair, pose, vibe)”
Now you have:
1 base character
5–10 variations
→ same identity, different “angles”
This is your UGC army
Step 3: Scale images with Arcads
Inside Arcads:
Create a workflow
Paste each JSON variant from Claude
Add a nanobanana image block (or any image gen model)
Duplicate block = number of variants
Now you’re generating:
→ multiple influencers
→ in parallel
→ fully automated
Step 4: Turn images → video
Create ONE master prompt:
“Generate a short-form video of this character posing / walking / talking with subtle motion, cinematic lighting, viral TikTok style”
Then:
Connect all image outputs
→ into 1 AI video generation block
Now you get:
→ multiple videos
→ same format
→ different characters
Why this works
You’re not creating content
You’re building a system that prints content
Variants = more hooks
More hooks = more chances to go viral
More viral = more installs / revenue
Real insight most people miss:
1 influencer = 1 chance
10 variants = 10 chances
Distribution beats perfection.Alex NguyenApr 26125301A repeatable playbook for B2C appsguideB2C App Playbook (2026):
1. Ship fast → weeks, not months. Ugly > perfect.
2. Pick ONE channel → TikTok, X, or Reddit.
3. Consistency > luck → 100 posts beats 1 viral spike.
4. Charge early → $5/month > 5,000 free users.
5. Retention compounds → growth hacks don’t.
6. iOS + Web first → that’s where the money is.
7. Win organic first → paid comes later.
Simple.
Brutal.
Repeatable.
If you want it edgier or shorter for max reach, say the word and I’ll tune it.Alex NguyenJan 26279258Improve proven products instead of copying blindlyinsightMust-read if you’re trying to build a successful app.
I really like this part: “Stop avoiding copying what works from other people. Doing this is actually wise product thinking.”
Because if all you do is copy another app step by step the product, features, and even their marketing strategy you’re basically building for them and helping market their product.
Over time, your product thinking will die. You lose creativity, become dependent on what others are doing, and will always be one step behind.
Instead, learn from what they’re doing right, understand why it works, and apply those lessons to your own product in your own way.
Learn from what works. Don’t blindly copy it.Alex NguyenAug 26122249TikTok uses local signals to target viewersguideLesson 2: You need a VPN because TikTok’s algorithm first targets local viewers. Try checking your audience traffic in TikTok Studio to see it yourself.
If you want to target users in the US while you’re in Vietnam like me, a VPN or proxy is necessary. If you’re already in the US, you don’t need one unless you want to scale with 10–30–100 accounts, in which case you’ll need it to avoid spam bans.
As for automation, if your account is already trustworthy and has lot of followers, it can work. But if it’s a completely new account and you use automation tools right away, I’m not so sure.Alex NguyenMar 26179249A short onboarding hurt app conversioninsightMy first failure of 2026.
I was very confident about my AI calories app, but I made a mistake.
The onboarding was too short and the core features didn’t stand out enough compared to other calorie tracking apps.
I started organic TikTok marketing and it actually worked pretty well. Traffic was coming in.
But almost 100% of users started the trial and then canceled. Conversion = 0.
So I decided to stop marketing for now and focus on shipping v2.0.0 to make the product much better.
Sometimes the problem isn’t distribution.
It’s the product. 🥴Alex NguyenMar 26509224Daily posting is the simplest marketing strategyguideIf you don’t know how to market, do this:
> Post every day
> Use a clean, trustworthy profile photo
> Make your bio stupidly clear (what you do + who it’s for)
> Reply to 100+ people (real humans, not AI spam)
> Connect with new creators daily
You’ll learn 10× faster by doing than by overthinking.
Show up every day.
That’s the simplest marketing strategy on earth.Alex NguyenJan 26365207Human-like TikTok workflows can still go viralinsightAnother guy followed the free sauce I shared and went viral with 300K+ views after posting consistently for 45 days and 98% US viewers (higher than mine lol 😭). Congrats, bro.
Shadow-bans are a joke. If you’ve only posted for a few days and haven’t gotten high views, change your content format before abandoning the account.
As I mentioned in my previous post, TikTok increases the risk of bans when you automate and your posting behavior looks unnatural. But if your workflow behaves like a real human user, TikTok usually won’t ban you, even if you use a VPN or proxy.
That said, a bad IP that has been heavily spammed on TikTok is still not good.Alex NguyenJul 26124189Most B2C founders overthink marketing.guideMost B2C founders overthink marketing.
This is the path I’ve seen work again and again:
$0–$20k
→ Post organic content yourself
→ Build a small AI content factory
$20k–$100k
→ Hire UGC creators to scale
→ Still keep posting organic personally
$100k+
→ Turn on paid ads
→ Scale without artificial limits
Don’t rush ads.
Organic creates leverage.
Paid ads just amplify what’s already working.Alex NguyenFeb 26206180Production basics separate apps from demosguideEvery vibe coder should understand this (Part 2) 👇
Stuff that separates “it runs on my laptop” from “real product”:
Shipping & prod basics:
- dev vs staging vs prod ≠ same thing
- build passes locally ≠ works in prod
- logs > guessing (read them)
- crashes in prod are normal, silence is not
- rollbacks > panic fixes
- config changes can break everything
APIs & integrations:
- rate limits exist (you WILL hit them)
- API keys ≠ frontend code
- retries + timeouts matter
- 3rd-party services WILL go down
- mock data before real APIs
Frontend survival:
- loading states aren’t optional
- empty states are UX, not bugs
- errors should be user-friendly
- mobile first (yes, really)
- slow UI = users bounce
Founder mindset shift:
- perfection blocks shipping
- boring solutions make money
- speed > elegance (early on)
- users don’t care about your stack
- shipping > tweeting about shipping
You’re not learning to code.
You’re learning to finish things.Alex NguyenFeb 26153160Find a winning format before automating TikTokguideEveryone wants to automate TikTok.
But automation should come AFTER you find a winning format.
Especially if you’re targeting the US with VPN/proxy.
My rule:
Test manually → find repeatable views → break down why it works → turn it into a template → automate the boring parts.
Automation won’t fix a bad format.
It only scales what already works.Alex NguyenJun 26148146Vibe coders need basic production safeguardsguideEvery vibe coder needs to understand this 👇
Not advanced engineering.
Just the boring basics that save you from breaking everything.
Non-negotiables:
> .env = secrets live here
> Never push .env to GitHub
>npm run dev = local only
> package.json = scripts, dependencies, versions
> npm run build = what actually ships
> npm packages = install / remove / update carefully
> git add .env = instant regret
> localhost:3000 ≠ a real URL
> env vars ≠ hardcoded values
> terminal errors are not decoration
Nice-to-know stuff that saves days:
> npm i --legacy-peer-deps when packages fight
> .gitignore = files Git should ignore
> basic Git: clone, pull, commit, push
> DB schemas / SQL if your app stores data
> auth basics: users, sessions, tokens
> dev words: components, modals, cards, routes
> reading docs > randomly guessing
> asking AI clear questions > “fix this bro”
You don’t need to be a “real dev.”
You just need to understand enough to ship without nuking your own app.
Happy vibe coding ✌️Alex NguyenMay 2697126Diagnosing TikTok posts stuck at zero viewsguideIf your TikTok gets 0 views, it usually means the video/slideshow hasn’t been pushed to the For You feed yet, or your account/video is being restricted.
For a new account, 0 views for the first few minutes or hours isn’t a big deal.
But if it’s still at exactly 0 after 12–24 hours, something is probably wrong:
→ Video is still under review/processing
→ Content looks like spam, a reupload, or duplicate content
→ New account posting too fast / behaving like a bot
→ VPN, proxy, or IP changes too frequently
→ Too many accounts on the same device with similar behavior
→ Video is accidentally set to Only Me/Friends
→ Account is temporarily restricted
If only ONE of your accounts gets 0 views → probably an account/video issue.
If MULTIPLE accounts suddenly get 0 views → check your device, IP, and posting workflow first.
Don’t keep spamming posts. It usually makes things worse.Alex NguyenSep 2610393No posts found
Try a different phrase or clear your filters.
Couldn't load the archive
Check your connection and try again.