- The Runway Ventures
- Posts
- 🤯 How Replyr scaled to $180K ARR with ONE simple growth hack
🤯 How Replyr scaled to $180K ARR with ONE simple growth hack
AhBengGPT. NajibAI. SOHAI. The same playbook that went viral and funded Replyr.ai.
Admond Lee
October 05, 2025
Hey Founders,
👋🏻 Welcome to The Runway Ventures (🎙️Podcast Edition) — where I interview founders and deep dive into their stories to help you become the top 1% founder by learning from their startup journey and mistakes with actionable insights.
Today’s story is about how Dylan built and scaled Replyr to $180k ARR without spending a single cent on ads (using ONE simple growth hack). Let’s get to it! 🚀
Today at a Glance:
🫡 1 Founder → Dylan Tan (founder of Replyr.ai)
⚠️ 2 Mistakes → Too attached to startup ideas
🧠 3 Lessons Learned → Validate the pain & demand (not just the idea)
🔗 The Runway Insights → The CLEAREST path to building a $100k MRR mobile app
💰 Southeast Asia Funding Radar → Neptune Robotics raises $52M (Series B) to build underwater robots that replace divers by cleaning ship hulls 3-5 times faster, automatically and and safely
🫡 1 Founder: Dylan Tan (founder of Replyr)
🌟 The Highlights
🇸🇬 Founded by Dylan Tan in 2023, Replyr (Iterative S24), pronounced "replier", converts leads to customers on WhatsApp by deploying AI Agents that can dynamically chat with customers while interacting with other business tools.
The Problem — 😢 Businesses get flooded with leads and inquiries on WhatsApp, Instagram, and other chat apps — but staff can’t keep up.
Slow or missed responses mean lost sales.
Traditional chatbots don’t fix this because they can’t qualify leads, book appointments, or integrate with business tools.
The Solution — 🤖 Replyr deploys AI agents that chat with leads in real time, 24/7, and convert leads to customers.
These agents don’t just answer questions — they can qualify prospects, schedule appointments, follow up, accept payments, and update CRMs.
In short, Replyr turns chats into paying customers by automating the entire workflow, not just the conversation.
🫡 Founder Story
🛍️ Dylan started out in 2018 with Split, a “buy now, pay later” startup born from him literally lending out his personal credit card to help Malaysians afford flights — a quirky idea that grew into a regional BNPL platform backed by 500 Global and even landed him on Forbes 30 Under 30.
💵 When funding dried up and BNPL’s margins collapsed, Dylan pivoted to Sugar, a “save now, buy later” app that rewarded people for saving — but it flopped fast.
🤖 Then, almost accidentally, he built a viral chatbot called AhBengGPT during the early ChatGPT craze, which made him internet-famous and sparked his next idea. From that moment came Replyr.ai, where Dylan now builds AI agents that chat, qualify leads, and close sales on WhatsApp — turning his failed fintech lessons into a thriving AI startup backed by top investors and global recognition.
 Dylan Tan (founder of Replyr) | 🤯 Before pivoting to Replyr, Split onboarded over 800 merchants and 100,000+ users, processing over RM100 million (~US$23M) in transactions and generating around US$60,000 in monthly revenue at its peak. By late 2021, it felt like the sky was the limit – Split had serious momentum. |
🚀 They secured a term sheet for a $10 million debt facility to fuel lending, and were deep into due diligence for a major Series A equity round with a venture capital firm.
However, the celebration was short-lived. Just as Split hit its stride, cracks began to appear. Rising interest rates and a crowded BNPL field started squeezing their margins and investor enthusiasm.
🤦🏻♂️ Then came the gut punch — the lead VC abruptly pulled out of the funding deal in early 2022, and it was later found out that they didn’t actually have the money to invest.
🌋 The Lowlights
I’ve forgiven them now – I consider it a scar I had to suffer, so I remember what not to do.
⚠️ After 4 gruelling months of due diligence calls and paperwork, the supposed lead investor told him, “Sorry, we decided not to fund you” — out of the blue.
Only later did he learn the harsh truth — the VC had been stringing him along without actual funds in their coffers.
🫡 Suddenly, Split was on the brink of dying. Dylan halted new sales, shared the layoff news with his employees (and made sure they were taken care of), and negotiated with Split’s merchants to wind down service gradually rather than suddenly pulling the plug.
People don’t want to save money, they just want to spend money.
🍬 Dylan was exhausted and heartbroken watching his “baby” fall apart. To make matters worse, their pivot project, Sugar, wasn’t panning out either. Sugar was the Save-Now-Buy-Later app they launched in mid-2022 after shutting down Split – basically the opposite of BNPL, encouraging people to save up for purchases instead of going into debt.
🥕🥕🥕 The idea was noble. But in reality, nobody really cared. People want to spend money, not to save money. No matter how many cashback rewards they dangled, it was hard to change consumer behaviour. After pouring months into it, he made the call to shelve Sugar too.
These were truly dark days for Dylan. The emotional toll was intense. But Dylan learned something…
I had to learn the hard way that a startup is not my child… it’s a science lab. Most experiments fail, and that’s okay. You throw away the ones that don’t work and try something else until something works.
That mindset shift — treating startup life more like a scientist than a parent — is what Dylan carried forward as he prepared for the next chapter.
🚀 The Comeback
AhBengGPT
In late 2022, right as Sugar was fizzling, a new phenomenon burst onto the scene: ChatGPT. Generative AI was suddenly accessible, and Dylan’s inner hacker was intrigued.
😂 His first experiment? A silly chatbot game on Instagram that he built over a weekend using a no-code tool. Dylan called it “AhBengGPT”. It was basically an AI-powered role-play of bargaining — something every Malaysian/Singaporean kid who’s dealt with a shady phone seller could relate to.
🤑 Dylan posted it on a Facebook group and it just went viral overnight. Surprisingly, a few companies reached out and wanted Dylan to build chatbot for their businesses. And just like that, Dylan made his first $1 — Replyr was born.
After talking to more potential customers and iterating on the product, Dylan pivoted Replyr towards workflow automation on WhatsApp with AI agents to qualify a lead, answer their questions, collect necessary info, and even perform actions like scheduling appointments or initiating orders.
This time, it worked!
By mid-2025, Replyr crossed $180k ARR without spending a single cent on ads.
How?
NajibAI
😂😂 Dylan had the knack for creating buzz for Replyr in the news. After launching AhBengGPT, he launched NajibAI — an AI tool that helped Malaysians track their expenses on WhatsApp. Once again, it went viral in Malaysia, and Dylan found his inbox full of inquiries from potential customers who discovered Replyr through the story.
Scammers On Hold AI (SOHAI)
😂😂😂 In early 2025, Dylan pulled off his biggest viral hit with “Scammers On Hold AI”, or SOHAI (ask your Chinese friends if you don’t know what SOHAI means 🤣). It was an AI to waste scammers’ time on WhatsApp:
If a scammer contacts you, or you see an ad for a scam, give them SOHAI's phone number.
SOHAI will pretend to be Madam Soh — she will infinitely waste their time by giving excuses, digress, and act blur. If the scammer doesn't reply, SOHAI will follow up with the scammer to annoy them back.
Every minute of scammer time wasted is one less minute of them scamming someone else.
I mean.. who doesn’t hate scammers? SOHAI exploded in popularity. Big brands started calling. And Dylan landed Malaysia’s biggest telco as his customer.
📌 Here’s the TLDR of Replyr:
Replyr — WhatsApp AI for Clinics
2018 (Aug) – Dylan left his corporate job at Skyscanner to co-found Split, a Buy Now Pay Later startup for travel.
Split won the WiT Singapore Startup of the Year pitch competition, validating the concept.
🛍️ Riding the BNPL wave with Split
2019 — 💰 Raised seed investment from Entrepreneur First and 500 Global, gaining resources to grow.
2020 — COVID-19 decimated travel. Dylan pivoted Split from travel to e-commerce BNPL, onboarding 800+ online merchants and over 100,000 users in Malaysia/Singapore.
BNPL demand surged with lockdown online shopping.
Late 2021 — 📈 Processed more than RM100 million in transactions and hit about US$60k in monthly revenue.
Secured a term sheet for $10M in debt financing and entered due diligence for a Series A equity round – Split’s future looks bright.
Early 2022 — 📉 A prospective VC withdraws from Split’s Series A at the last minute, leaving the company without runway.
Facing market saturation and rising costs, Dylan made the painful decision to shut down Split. Half the team was laid off and operations wound down gracefully over months.
🥕 Save more with Sugar
Jun 2022 — Launched Sugar, a “Save Now, Buy Later” fintech app, hoping to encourage consumers to save for purchases (an idea spun out of user feedback from Split).
Early users signed up and Sugar operated with a handful of merchants.
Late 2022 — 🙅🏻♂️ Sugar failed to gain wide adoption as most shoppers preferred instant gratification over saving up.
Dylan observed that user behaviour won’t budge (people just want to spend, not save).
He decided to sunset Sugar and pivoted again. Around the same time, ChatGPT launched, sparking Dylan’s interest in AI.
🤖 AhBengGPT → NajibAI → SOHAI → Replyr
Jan 2023 — 😂 Experimenting with generative AI, Dylan built “AhBengGPT”, a humorous Instagram chatbot game that let users haggle with an AI persona.
It went viral overnight, unexpectedly kickstarting Dylan’s next startup idea.
Mar 2023 — Riding the momentum, Dylan officially founds Replyr to help businesses deploy AI chat agents.
Nov 2024 — 😂 Launched “NajibAI”, a free WhatsApp receipt-tracker bot. Its cheeky name and useful function grab media attention.
The publicity generated inbound leads and established Replyr’s reputation in AI solutions. By this time, Replyr has grown to about $100k ARR purely via organic growth.
Feb 2025 — 😂 Dylan launched “Scammers On Hold AI (SOHAI)”, a WhatsApp bot that wastes scammers’ time.
SOHAI exploded in popularity, with coverage by major news outlets.
Mid 2025 — 🔥 Replyr’s successes led to big opportunities.
Dylan won a global travel tech pitch-in Barcelona, sparking partnerships in the travel industry.
Large brands (including Malaysia’s top telecommunications company) signed on as clients, impressed by Replyr’s unique track record.
The startup reached around $180k ARR and is steadily growing, marking an incredible comeback from the brink of failure.
📈📉📈 Dylan went from the lowest low — watching his fintech startup flame out — to a roaring high, building an AI venture that’s making waves. And he did it by staying scrappy, listening to the market, and not being afraid to launch weird and wonderful ideas to see what sticks.
Dylan’s story shows that true founders don’t quit. They adapt, they experiment, and they come back stronger.
Want to learn more about Replyr?
⚠️ 2 Mistakes
Mistake 1: Too attached to startup ideas
🙏🏻 In Dylan’s first startup (Split), he was deeply attached to the idea — emotionally and personally, realising how unhealthy that mindset had been.
Back then, he poured everything into making his original vision work — even when market signals (like COVID decimating travel) screamed otherwise.
He refused to pivot early, holding onto the product for too long because it felt like giving up on himself.
Mistake 2: Built products before validating demand
We designed a nice product, a rewarding product… but people don’t want to save money, they just want to spend money.
When Dylan Tan built Sugar — his “Save Now, Buy Later” app — it wasn’t a bad idea on paper.
It even sounded noble — help people save first, then buy later.
🙏🏻 But Dylan made the same classic mistake many founders (especially product-driven ones) fall into — he built a “smart” product before confirming if anyone actually wanted it. Dylan built based on what he wished was true, not what users actually did.
He was solving a psychological problem (financial discipline) instead of a business pain. Users didn’t care about responsible spending—they cared about instant gratification.
The result? Zero traction despite a beautiful product and a clear mission.
🧠 3 Lessons Learned
Lesson 1: Treat startup ideas as experiments (not your child)
Your startup is not you – it’s a lab. Most experiments fail, and that’s fine. Learn from it and double down on what works.
Back then, every decision felt personal.
A product feature failing meant he failed.
A pivot felt like a betrayal of his vision.
He poured everything into proving he was right — instead of testing if he was.
… that attachment blinded him to what the market was really saying.
In AI startups, this is gold. Models, prompts, and use cases evolve at lightning speed.
The winners aren’t those who “believe hardest” in their idea — they’re those who ship experiments, gather data, and pivot on insights.
🌮 Key Takeaways:
🙏🏻 Detach emotionally from outcomes
Your worth isn’t tied to your product. If an experiment fails, you didn’t fail — the hypothesis did.
🧪 Run fast & cheap experiments
Before you hire engineers or raise money, validate your assumptions with no-code tools, mockups, or even manual operations.
Dylan’s first paying Replyr client came from a scrappy, hacked-together chatbot that barely worked — but it proved willingness to pay.
In fact, you can build a fully functional MVP with Lovable or Bolt, show it to your users, get feedback, get paid, and iterate.
You’re literally a few prompts away from building a working product for your users. This is the BEST time to build now (seriously).
✍🏻 Document everything
Keep an “experiment log” — what you tested, what you expected, what actually happened. This is how you find signal in chaos.
Look at the data, iterate and pivot accordingly.
Lesson 2: Validate the pain & demand (not just the idea)
When Dylan started Replyr, he flipped his process. He didn’t build what he thought businesses needed — he watched what they asked for.
One client requested an AI that could handle WhatsApp leads became a validated pain point, which grew into Replyr’s core product.
🌮 Key Takeaways:
🩺 Start with symptoms (not solutions)
Talk to users and look for frustration — what are they doing manually, repeatedly, or resentfully? That’s where pain hides.
Dylan found that businesses hated manually following up with WhatsApp leads — that was the real pain.
Here are some questions you can ask when doing user interviews to uncover what users “actually” want (not “I think I’ll like it” or “this sounds interesting”):
Do you face this problem?
Can you walk me through the last time you faced this problem?
Why was it hard?
Have you tried any solutions to solve this problem?
Why don’t you love about the solutions you’ve tried?
🧐 Ask “How are you solving this now?”
If users are already hacking together Google Sheets, freelancers, or WhatsApp templates to solve it — bingo, that’s validation.
If they’re doing nothing, the pain isn’t strong enough yet.
🤑 Charge early, even for an MVP.
Dylan’s first bot was “crappy” by his own words, but it brought in a paying customer.
Real demand is measured in dollars, not likes.
The biggest validation is when someone pays you to solve their problems.
Lesson 3: Choose investors like co-founders
During Split, Dylan was constantly fundraising. As a first-time founder, every “yes” from an investor felt like validation. But that rush blinded him to alignment issues.
When one VC later pulled out of a Series A deal because they didn’t actually have funds, the fallout nearly killed his startup.
That’s why for Replyr, he deliberately kept the cap table lean, only bringing in Iterative who matched his growth ethos, and avoided “tourist VCs” chasing hype cycles.
🌮 Key Takeaways:
🧠 Test for Philosophy Fit (not just Term Sheet Fit)
Ask investors how they handle failures in their portfolio. If they flinch or sound transactional, run.
You want people who’ll sit with you in the mud, not just toast you in the spotlight.
✊🏻 Check their “Aftercare”
Founders often check an investor’s network but forget to ask, “What happens when things go south?”
Talk to founders who’ve failed under that fund — how did the investors react?
🧐 Watch their questions.
Smart, aligned investors ask, “How are you thinking about X?”
Misaligned ones ask, “When do I get my 10X?” — before you’ve even found product-market fit.
🤝🏻 Pick those who back humans (not headlines)
Dylan’s best backer (Iterative) didn’t invest because Replyr was “the next ChatGPT”.
They invested because they believed Dylan could learn fast and adapt.
That’s the kind of investor who compounds over time.
🔗 The Runway Insights
The CLEAREST path to building a $100k MRR mobile app (Link)
Handling multiple ICPs as a solo founder (Link)
5 interesting learnings from Figma at $1 billion ARR (Link)
A 4-step framework for building delightful products (Link)
The right way to pitch VCs and accelerators (and why most founders get It wrong) (Link)
💰 Southeast Asia Funding Radar
Neptune Robotics raises $52M (Series B) to build underwater robots that replace divers by cleaning ship hulls 3-5 times faster, automatically and and safely (Link)
Honest raises total equity to $100M to expand its credit card offerings (Link)
SwarmFarm bags $19.8M (Series B) to build self-driving robots for agricultural use (i.e. autonomous SwarmBots) (Link)
Art of Time raises $19.5M (Series B) to sell luxury watches through its website and stores (Link)
🤝🏻 Before you go: Here are 2 ways I can help you
Founder Office Hours: Book a 1-1 call with me, share your problems and questions, and I'll help you cut through the noise, avoid costly mistakes, and get clear next steps that actually work. I help early-stage founders with:
Validating ideas & building MVPs
Tech & product development
GTM strategy & fundraising
Finding PMF & growth hacks
Growing & monetising newsletters
Attract customers & investors by building a solid founder brand on LinkedIn
Promote your business to 18,000+ founders: Acquire high-value leads and customers for your business by getting your brand in front of highly engaged startup founders and operators in Asia.
💃 Rate Today’s Edition
What'd you think of today's edition?Your feedback helps me create better content for you! |
That’s all for today
Thanks for reading. I hope you enjoyed today's issue. More than that, I hope you’ve learned some actionable tips to build and grow your business.
You can always write to me by simply replying to this newsletter and we can chat.
See you again next week.
- Admond
Disclaimer: The Runway Ventures content is for informational purposes only. Unless otherwise stated, any opinions expressed above belong solely to the author.
Reply