🤯 NocNoc Shut Down After ฿4.39B in 5 Years of Losses

New here? Join us to become the top 1% founder by learning from startup failures:

🤝🏻 You agree a price in US dollars with your supplier and pay 60 days later.

If the US dollar rises 2% by then, the same invoice costs you 2% more in Singapore dollars or ringgit. If your margin on that order is 10%, that's nearly 20% of your profit gone. And that's before your bank's own markup on the rate, which never shows up as a fee.

💸 Nothing about the order changed, but you made less on it.

With WorldFirst, a brand of Ant International (same family as Alipay+), you can convert to US dollars when the rate suits you, or set a target rate and let it convert automatically, 24/7. Then hold the dollars until the invoice is due. You see the rate and fees before you confirm, so you know the full cost upfront.

PRISM+, the Singapore electronics brand, saves five figures a month with WorldFirst.

Start protecting your margin →

P.S. Get a US$100 bonus on your first US$1,000 transaction. Just sign up before 31 December 2026 and enter EXCLUSIVEOFFER in the Invitation Code field.

Today at a Glance:

• ☠️ 1 Failed Startup → NocNoc

• ⚠️ 2 Mistakes → Running a full-stack service business on marketplace commissions

• 🧠 3 Lessons Learned → Your investors' patience sets your runway

• 🔗 The Runway Insights → Will agents have network effects?

• 💰 Southeast Asia Funding Radar → Manus raises $500M+ after China blocked Meta's $2B deal

🚪 Knock knock, SCG's at the door

• Feb 2019 — 🚀 NocNoc soft-launched as a one-stop home-and-living marketplace with a ~US$6 million build budget.

• Jun 2019 — Operating in 35 provinces through 52 partner installer crews, with ~120,000 visits a month.

  • But first-year numbers already told the story: THB 5 million revenue vs THB 155 million loss.

• 31 May 2023 — 💰 Must Be Company (ThaiBev + Frasers Property) subscribed THB 1.953 billion for 50% of BetterBe, NocNoc's operator.

• 24 Oct 2023 — AWS case study revealed NocNoc had 16.6 million users and had more than doubled sales using machine learning.

• 25 Jan 2024 — 🎯 NocNoc set an ambitious THB 6.5 billion sales target for 2024, claiming 600,000+ products and 6,000 sellers.

  • NocNoc never published its 2024 sales. BetterBe's filings show its revenue fell 22% that year, to THB 322 million.

🚪🔒 Nobody's home anymore

• 8 Nov 2024 — ⚠️ Must Be quietly made available a THB 1.15 billion shareholder loan facility to BetterBe.

  • Frasers only disclosed it 13 months later, in a filing.

• 1 Jul 2025 — 🔁 NocNoc announced Anupong Tasaduak (ex-MD of Renos) would take over as CEO from founder Chonlak.

• 25 Dec 2025 — Frasers Property signed up to put up to THB 938.6 million more in through Must Be, and SCG was expected to add its share.

• 9 Jan 2026 — 📢 Just 2 weeks later, BetterBe told Thailand's e-commerce regulator (ETDA) it was shutting NocNoc.

  • SCG's filing called it a joint decision of the shareholders, and said SCG expected to book a ~THB 1.8 billion non-cash loss for it in Q4 2025.

• 9 Feb 2026 — 🛑 NocNoc stopped accepting new orders through all channels at 00:00.

• 8 Apr 2026 — Final deadline for pending deliveries; general after-sales ended at 18:00.

• 9 May 2026 — ⚰️ NocNoc ended all platform services at 23:59.

🌮 Key Takeaways:

• Your take rate must cover the real cost of what you deliver, not the benchmark of the category you want to be compared with.

• If gross margin cannot cover coordination costs, scaling turns a small per-order loss into a bigger per-order loss.

🛠️ Operator Playbook:

• 🧮 Run a per-order P&L on your top 100 orders

  • Columns: revenue, commission, delivery, installer/service, support, warranty reserve.

  • If contribution per order is negative, you have a pricing problem — not a scale problem.

• 📊 Benchmark against who actually does your job

  • Compare your take rate with what a full-service contractor charges for the same job, not with horizontal marketplaces.

  • List the costs that contractor's price already covers — shipping, installation, warranty — and check which of them your fee actually pays for.

  • If one fee can't carry the service, add a second revenue line — installation fee, service plan, premium tier.

🌮 Key Takeaways:

• Quarterly profit cycles compound into founder deadlines, even when the balance sheet looks limitless.

• Decade-long bets need decade-long investors. Anything shorter is a hidden deadline on your cap table.

🛠️ Operator Playbook:

• 🗺️ Build a cap-table patience matrix before you raise

  • For each investor, map 3 columns: fund life remaining (or parent-co profit-reporting cycle), public-market reporting frequency, and reserved follow-on capacity.

  • Compare each row against your projected years-to-breakeven; flag any gap >3 years as a signalling risk and either renegotiate reserves or line up a bridge investor with a longer horizon.

  • Shopee and Lazada each ran 10+ years before turning a profit. Ask whether your backers could fund that wait.

🌮 Key Takeaways:

• Set next year's target on contribution per order, not sales. Because you can still hit a sales target while losing more on every order.

🛠️ Operator Playbook:

• 🩺 Diagnose the signal before deploying fresh capital

  • If losses grew faster than revenue last year, that is a model signal, not a scale signal.

  • Ask: "What specifically changes about per-unit economics if we spend this on scope versus iteration?"

  • If the answer is "nothing", you are funding expansion into a leak.

• 🧪 Pilot new categories as revenue tests, not expansion bets

  • Launch 1 new category at <5% of GMV before adding more.

  • Measure contribution margin, repeat rate, support cost, return rate — against the existing core.

  • Only roll out when the pilot beats the core on at least 2 of those 4 metrics.

🔗 The Runway Insights

• Will agents have network effects? (Read)

• The AI monetisation debate shifts again (Read)

• Prediction: the next massive Google update is around the corner (Read)

• $20M from Accel & Index to replace Slack — with no co-founder (Read)

• We're building multiplayer AI. Here's what we've learned so far (Read)

• Harvey vs Legora: when are terrible gross margins OK? (Read)

💰 Southeast Asia Funding Radar

• Manus raises $500M+ after China blocked Meta's $2B deal (More)

• OneByZero secures $20M (Series A) to expand AI-driven data and analytics solutions across enterprise clients regionally (More)

• Boost, Axiata's Malaysian fintech arm, gets $20M in equity from IFC at a $340M valuation to grow its digital lending business (More)

• 3cat bags $4M (Series A) to expand its pre-owned electronics marketplace from Malaysia into the Philippines (More)

• Zeya Health acquires ConnectLah to build AI infrastructure for healthcare clinics across the Southeast Asia region (More)