SureMine — reservation gate for Philippine live sellers

1. One-liner

Turns a “mine” comment into a held downpayment and screens repeat ghosters for Filipino live sellers.

2. Trend signal — why now?

Filipino live selling runs on one ritual: a viewer types “mine” or “pa-mine po” on a Facebook or TikTok live to claim an item, then is supposed to DM the seller and pay. The problem the whole market complains about by name: the “joy reserver” / “joy miner” / “bogus buyer” — someone who claims the item, locks it away from real buyers, then ghosts. Sellers describe it bluntly: buyers who “say ‘mine’ to the items but when it’s time to pay, pick-up or delivery, they will not communicate anymore and ignore your messages” — a “total waste of time and effort.” There is a Change.org petition demanding bogus buyers face jail, dedicated Facebook albums naming and shaming them, and a viral run of TikTok videos from lawyer Atty. Chel Diokno explaining that a “mine” comment is a legally binding contract of sale under Philippine law.

The money side is worse than the wasted time. ~60–68% of PH online orders are still COD, and COD return-to-origin (RTO) runs 20–40% — the seller eats both the outbound and return shipping when a buyer ghosts. So every joy-miner isn’t just lost time; it’s a negative-margin shipment.

What changed in the last 12 months: (1) PayMongo / Maya / Xendit now expose dead-simple payment-link and e-wallet APIs (GCash + Maya cover ~90% of PH wallet users; GCash alone has 94M registered), so collecting a ₱50–200 downpayment the second someone says “mine” is finally frictionless; (2) cheap multilingual LLM inference can read live comment streams in Taglish and auto-draft the confirm-or-cancel chase; (3) PH social commerce hit $28.4B in 2025, growing ~14.5% CAGR — a big, fast-moving pool of sellers feeling this pain weekly.

Provenance:

• Signal 1 (demand): Joy-reserver / bogus-buyer epidemic — named, petitioned, lawyered. “say ‘mine’… then will not communicate anymore” — https://read.cash/@MizLhaine/bogus-buyers-joy-reservers-problems-in-online-selling-e02a815d ; Change.org petition https://www.change.org/p/lazada-bogus-buyer-should-face-imprisonment ; Atty. Chel Diokno “pa-mine” warning https://interaksyon.philstar.com/trends-spotlights/2022/11/11/234336/pa-mine-po-chel-diokno-cautions-joy-miners-in-live-selling-sessions/ — observed 2026-06-19
• Signal 2 (feasibility): PayMongo/Maya/GCash payment-link & e-wallet APIs now trivial for micro-sellers; e-wallets cover ~90% of PH digital payers — https://hitpayapp.com/blog/ecommerce-payment-gateways-philippines , https://webdesigner.ph/articles/gcash-maya-paymongo-philippine-payment-integration-guide/ — observed 2026-06-19
• Signal 3 (economic): PH social commerce $28.4B in 2025, ~14.5% CAGR; COD RTO 20–40% bleeding seller margin — https://www.imarcgroup.com/philippines-social-commerce-market , https://www.cloudecommerce.com/blog/how-to-reduce-cod-rto-rates-in-the-philippines-7-proven-strategies-for-2026/ — observed 2026-06-19 Category: Geographic arbitrage

3. The opportunity

The incumbent is Page365 — 350K PH sellers, auto-creates an invoice the moment someone types “MINE.” Good. But it solves the clerical half of the problem (turning a comment into an order) and leaves the expensive half wide open: it does nothing to make the buyer actually pay. The invoice still goes to a stranger who has risked nothing. CommentSold and SoldLive (US/Shopify) have the same blind spot and aren’t localized for GCash/COD anyway.

The gap: there is no commitment layer. No product turns the “mine” moment into a small held downpayment, screens the commenter against a shared history of ghosting, and runs the confirm-or-release chase automatically. An AI-first PH-native team can own that layer. The wedge is brutally simple — make the joy-miner put ₱100 down before the item is held for them — and the AI does the work the seller currently does by hand at midnight: read every “mine,” match it to a buyer, fire the payment link, chase the silent ones, release unpaid holds back to the live, and update a cross-seller reputation score.

4. Target market

• Primary customer: Full-time and side-hustle Facebook/TikTok live sellers in the Philippines — solo operators and 2–5 person “online shop” teams selling apparel (ukay/thrift), preloved goods, beauty, toys, and household items. They run 3–7 lives a week, 50–500 viewers each, ₱30K–₱300K monthly GMV.
• Why they buy (their words): Joy reservers are “a total waste of time and effort”; they “reserve the item, thus depriving other buyers” and then ghost, so “the online seller bears the losses.” On COD the loss is real cash — double shipping on every RTO. They already manually demand downpayments via GCash screenshots and keep private “bogus buyer” lists; they want that automated and shared.
• Rough TAM reasoning: PH social commerce is $28.4B (2025). Hundreds of thousands of active live sellers (Page365 alone has 350K seller accounts). Capturing even 20–40K paying live sellers at ₱500–1,500/mo is a ₱120M–₱720M/yr (~$2M–$12M) revenue band — comfortably inside the sub-$5M ARR target on the low end, with headroom.
• Why now for them: GCash/Maya links made instant micro-downpayments frictionless in the last year; COD RTO pain is rising as volumes grow; and the joy-miner discourse is at a cultural peak (petitions, viral lawyer clips). Sellers are primed to pay to make it stop.

5. Product sketch (MVP)

• “Mine” capture: Connect a Facebook Page / TikTok Shop; SureMine watches the live comment stream and detects “mine / pa-mine / +1 / [item code]” claims in Taglish, matching each to the item the host is currently showing.
• Instant downpayment link: The moment someone claims an item, auto-DM them a GCash/Maya/PayMongo link for a seller-set downpayment (e.g. ₱50–₱200 or % of price). Item is “held” only once paid.
• Auto-release timer: Unpaid holds expire after a seller-set window (e.g. 10 min) and the item is announced back as available in the live — no more dead reservations.
• Ghoster reputation score: Every buyer gets a cross-seller SureScore based on paid-vs-ghosted history. New claimants flagged “high-risk” can be required to pre-pay before the hold is granted.
• AI confirm-or-cancel chase: For sellers who still allow COD, an AI agent DMs a polite Taglish confirmation (“Confirm po ba kayo? Reply YES para i-ship”) and only ships on a clear yes — killing the RTO at the source.
• Post-live order summary: Clean tally of paid / held / released / shipped, exportable, with the day’s recovered-revenue number front and center.
• One-tap blocklist: Seller marks a buyer bogus; it feeds the shared reputation graph (privacy-safe — score, not public shaming).

6. AI angle — what’s load-bearing

AI does three jobs a rules engine can’t: (1) parses messy Taglish live comments — “mine yung pink size M”, “+1 sa floral”, “akin na po yan” — into structured claims matched to the on-screen item, in real time, across hundreds of comments a minute; (2) runs the confirm-or-cancel conversation in natural Taglish, handling “san pickup?”, “pwede COD?”, “sched lang muna” and deciding whether that’s a real commitment or a soft ghost; (3) scores buyer intent from behavioral history to set who must prepay. Remove the AI and you’re back to the seller squinting at a comment wall at midnight and typing replies by hand — which is exactly the status quo. The AI is the product, not a sticker.

7. Localization angle

This is a localization play end to end and cannot be a generic global product:

• Payment rails: GCash + Maya + PayMongo links are the mechanism. Stripe/PayPal are irrelevant here.
• Language: Taglish comment parsing and chase dialogue. A US-English tool misreads the entire comment stream.
• The “mine” ritual + COD reality: The whole product is shaped around the PH “pa-mine” reservation culture and 60–68% COD share. SoldLive/CommentSold assume prepaid Shopify checkout — a different planet.
• Pricing: ₱499–₱1,499/mo tiers fit PH micro-seller wallets where a $49/mo tool can’t.
• Distribution: Filipino FB seller groups and the joy-miner shaming community are the channel. Expansion to Indonesia (Shopee/TikTok live, similar “keep” ritual, similar COD RTO) and Thailand is the obvious sequel — but PH first.

8. Business model — path to $1M–$5M ARR

• Pricing: Tiered SaaS. Starter ₱499/mo (1 connected page, downpayment links, basic tally). Pro ₱999/mo (reputation scoring, AI confirm-chase, auto-release, multi-platform). Shop ₱1,499/mo (multi-seller team, analytics). Optional 0.5–1% take on collected downpayments as a usage upsell.
• ACV: ₱9,000–₱14,000/yr ($160–$250) blended.
• Math to $1M ARR (~₱58M): ~5,000 sellers on the ₱999 Pro tier × 12 ≈ ₱60M ≈ $1.06M. Against 350K+ addressable sellers, that’s ~1.4% penetration.
• Math to $5M ARR: ~22,000 paying sellers blended at ~₱13K ACV, or 12,000 sellers + meaningful downpayment take-rate revenue. Requires winning the live-seller mindshare race against Page365 and surviving the Meta API dependency (see risks).
• Expansion path: ACV grows via (a) downpayment take-rate as GMV scales, (b) per-extra-page seats, (c) a “verified buyer” consumer side later (buyers pay to carry a trusted SureScore and skip prepay), (d) Indonesia/Thailand geographic expansion.

9. Go-to-market wedge — first 100 customers

• Raid the joy-miner communities. There are public Facebook groups and “bogus buyer” albums where sellers vent and name ghosters daily. Post in the 10–20 largest PH online-seller groups with a 30-second clip: “Stop joy reservers — make them put ₱100 down before you hold the item.” These people are already organized around this exact pain.
• Hijack the lawyer/discourse moment. Atty. Chel Diokno’s “pa-mine = contract” clips have huge reach. Partner with 3–5 PH selling-tips creators / “online shop ate” influencers for paid walkthroughs — their audience IS the customer.
• Live-watch outreach. Scrape currently-live PH selling pages, watch for hosts manually begging “please pay your downpayment within 10 minutes,” DM the seller a Loom showing SureMine doing it automatically. Highly targeted, expect strong reply rates from sellers mid-pain.
• Page365 switch/complement angle. Target sellers already paying for invoicing tools who still get ghosted: position SureMine as the missing “make them pay” layer. Comparison landing page.
• Free SureScore lookup. Public “check this buyer” tool (enter buyer’s FB handle, see ghost-risk) as a viral lead magnet that seeds the reputation graph.

10. Build complexity — justification

Medium. Payment links (PayMongo/Maya/Xendit), Taglish comment parsing, and the chase agent are all off-the-shelf-API work. The gnarly parts are (1) reliably ingesting live FB/TikTok comment streams within platform API limits and (2) building the cross-seller reputation graph cleanly (privacy, dispute handling). A focused 2–3 person team ships a credible v1 (FB-only, GCash downpayment + auto-release + basic tally) in ~3–4 months, with reputation scoring and TikTok following.

11. Gating checklist

Gate	Pass?	Note
Legal in target market	✅	Downpayments and “mine”-as-contract are explicitly recognized in PH; reputation scoring must avoid public shaming/libel (score, not names).
Ethical — no harm / dark patterns	✅	Reduces fraud and protects buyers from non-delivery too; blocklist is risk-score, not a public shame wall.
Market exists (evidence above)	✅	$28.4B market, named epidemic, petitions, 350K-seller incumbent.
1–5 person team can build this	✅	2–3 people, ~3–4 months to v1.
Launchable with <$50K / ₱40L	✅	API-cost light; main spend is the FB seller-group/influencer GTM.

12. Feasibility score

Axis	Weight	Score	Notes
Problem intensity	20	18/20	Money + time lost on every live; culturally named, petitioned, lawyered. Hair-on-fire.
Demand evidence	15	12/15	Multiple independent signals (epidemic discourse, petition, $28B market, COD RTO data). Gap: no proof yet that sellers will pay for this specific fix vs. their manual GCash-screenshot workaround.
Build feasibility	15	11/15	Mostly off-the-shelf APIs; FB/TikTok live-comment ingestion within platform limits is the real engineering risk.
Distribution clarity	15	13/15	Named FB seller groups, joy-miner albums, lawyer/creator audience. First 100 is a 2–3 week sprint.
Revenue mechanics	15	12/15	₱ tiers fit wallets, clear WTP vs Page365, optional downpayment take-rate. ACV modest, so needs volume.
Time to first revenue	10	8/10	Self-serve to micro-sellers; pilot-to-paid in weeks. Connecting FB pages adds a little onboarding friction.
Defensibility	10	7/10	Cross-seller reputation graph compounds with usage — real soft moat. Comment-capture itself is copyable.
Total	100	81/100

13. Qualitative modifiers

Founder-fit tags

technical-heavy (real-time comment ingestion, payment integrations, scoring) · operations-heavy (PH seller-community GTM, dispute handling on the blocklist)

Key assumptions to validate (3–5)

1. Assumption: Live sellers will pay ₱500–1,500/mo to cut joy-miners and COD RTO. How to test: 30–40 seller interviews in the top FB seller groups + a pre-sale landing page; target ≥10% “take my money” / deposit-for-pilot.
2. Assumption: FB/TikTok platform APIs allow reliable real-time ingestion of live-comment streams at the volume needed. How to test: Build the ingestion spike against a live test page in week 1; confirm latency and rate limits before anything else.
3. Assumption: Buyers will actually tolerate a ₱50–200 downpayment-to-hold without killing the seller’s conversion. How to test: Run it on 5 pilot sellers’ lives for 2 weeks; measure claim→pay conversion vs. their prior ghost rate. Sellers must net out ahead.
4. Assumption: A cross-seller reputation score is legally safe (not libel) and accurate enough to act on. How to test: Legal review on score-vs-shaming; backtest score accuracy against pilot sellers’ private bogus-buyer lists.

Risk flags

1. Platform dependency: Hard reliance on Meta/TikTok comment APIs and policy. They can throttle, change terms, or ship a native “deposit to reserve” feature. Mitigate by owning the payment + reputation layer (stickier than comment-reading) and going multi-platform early.
2. Two-sided friction: Downpayments could depress claim volume if sellers over-tune them; the product must prove net revenue gain, not just fewer ghosts. Default settings matter a lot.
3. Reputation-graph cold start + abuse: Score is weak until enough sellers feed it, and false “bogus” tags could harm real buyers. Needs dispute flow and conservative thresholds from day one.
4. Incumbent fast-follow: Page365 (350K sellers) could bolt on a downpayment feature. Speed and owning the cross-seller reputation data are the defense.

14. Structured verdict

Score:                  81/100
Verdict:                STRONG GO
Confidence:             Medium
Best-fit builder:       Technical founder (real-time + payments) paired with a PH live-selling-community operator
Time to revenue:        4–8 weeks from MVP (pilot-to-paid with seller groups)
Capital to launch:      ₱1.5M–₱3M ($25K–$50K) — mostly GTM and a small build team
Top 3 assumptions to validate first:
  1. Sellers pay ₱500–1,500/mo for it — 30–40 interviews + pre-sale landing, target ≥10% commit
  2. FB/TikTok live-comment ingestion is technically reliable at volume — week-1 spike
  3. Downpayment-to-hold raises NET revenue per live — 2-week pilot on 5 sellers, measure claim→pay vs prior ghost rate
Kill criteria:
  - Abandon if <10% of 40 interviewed sellers will pre-pay or pilot
  - Abandon if pilot sellers see claim→pay conversion fall enough to net LESS revenue than COD-with-ghosting
  - Abandon if Meta/TikTok ships native deposit-to-reserve before v1, or blocks live-comment ingestion

15. Next step — 1-week validation sprint

• Day 1–2: Build the FB live-comment ingestion + “mine” parser spike against one test page. Falsifiable: can we reliably capture and structure claims in real time within API limits? If no → hard stop.
• Day 3–4: Interview 30–40 active live sellers from the top PH seller FB groups. Show a clickable mock (claim → auto GCash link → auto-release). Ask for a ₱500 pilot deposit. Falsifiable: ≥10% put money down or commit to pilot.
• Day 5: Go / no-go. Go only if (a) ingestion works in real time AND (b) ≥10% of interviewed sellers commit cash/pilot. Anything less → revisit the wedge (maybe pure reputation-lookup as a thinner first cut) before building.