SB StartupBasket
All ideas
75 /100 GO Medium complexity

GreenLight — deliverability guard for WhatsApp agencies

Watches every WhatsApp number you manage and catches quality drops, template pauses, and bot-ban risk before Meta restricts you.

DevTools SaaS India+Global WhatsApp-firstCompliance-drivenAI-agentSMBSolo-builderPlatform-shift
— views
Evaluation Scores
75/100

GO

Overall Score

16
Problem
12
Demand
11
Build
11
Distrib.
12
Revenue
8
Time
5
Defense

GreenLight — deliverability guard for WhatsApp agencies

1. One-liner

Watches every WhatsApp number you manage and catches quality drops, template pauses, and bot-ban risk before Meta restricts you.

2. Trend signal — why now?

Meta rewrote the rules of the WhatsApp Business Platform in the last eight months, and the changes all point the same direction: tighter, automated, no-appeal enforcement that can kill a number overnight.

  • The general-purpose chatbot ban took effect for all existing accounts on 15 January 2026 (new accounts since 15 Oct 2025). Open-domain, ChatGPT-style bots are now prohibited; only “task-specific” bots (orders, bookings, support, notifications) are allowed. The line between the two is judged by Meta, not defined by you.
  • Unanswered-message tracking (2026): Meta now counts messages that get no reply within 48 hours. Cross a threshold and your messaging capability gets throttled. Low engagement is now a punishable offense, not just a vanity metric.
  • Quality ratings (green/yellow/red) + template pausing: templates auto-pause on negative feedback or low read-rates (error 132015), and a red number is blocked from scaling to the next tier. Tier upgrades are now checked every 6 hours.
  • Portfolio-shared limits (Oct 2025): all numbers in the same Business Manager share the highest achieved tier — so one bad number can drag, or one ban can endanger, the whole portfolio.

The buyers most exposed: the hundreds of thousands of SMBs and agencies on BSPs like AiSensy (210K+ businesses), Gupshup/Engati, Wati, and Interakt. BSP dashboards show quality after a template is already paused. Nobody pre-flights the risk or watches the chatbot-ban line.

Provenance:

3. The opportunity

Meta turned the WhatsApp Business Platform into a black box that silently punishes you. You learn your template is paused, your quality is red, or your bot tripped the general-purpose ban after the damage is done — Meta emails you, your delivery is already throttled, and there’s no real appeal. For a marketing agency running 5–50 client numbers, that’s a recurring fire drill: an angry client, a dead campaign, and sometimes a lost retainer.

The incumbents won’t fix this. BSPs (AiSensy, Wati, Interakt, Gupshup) each show a quality dashboard, but: (a) it’s reactive — it reports the pause, it doesn’t predict it; (b) it’s siloed per BSP, so an agency juggling clients across three providers has three dashboards and no portfolio view; (c) they have zero incentive to flag their own deliverability failures or tell you your bot is drifting toward a ban, because that’s an admission their platform let you do it.

GreenLight is the independent, cross-BSP layer that sits on top of every number you run, scores risk before you send, and warns you the moment quality slips — so you fix it before Meta restricts you. It’s a smoke detector for the one asset an agency can’t afford to lose: the number.

4. Target market

  • Primary customer: WhatsApp marketing/automation agencies and freelance “WhatsApp consultants” — typically 1–10 people — who manage 5–50 client WhatsApp Business numbers as a reseller under a BSP. Secondary: in-house growth marketers at D2C brands in India / SEA / LATAM running their own high-volume WhatsApp marketing (10K+ contacts).
  • Why they buy (in their words): “Broadcasting to dead numbers tanks your quality rating faster than anything — monitor your quality dashboard closely during the first few weeks on a new BSP.” They’ve felt the pain: utility templates silently reclassified to marketing (killing delivery), templates paused mid-campaign, support replies taking 2–3 days. A banned client number is a fired agency.
  • Rough TAM reasoning: AiSensy alone reports 210K+ businesses; YCloud 200K+; add Gupshup, Wati, Interakt, Twilio, 360dialog, and dozens of regional BSPs and the addressable base of API-active senders is well into the millions globally. Even capturing the agencies/resellers — a few hundred thousand operators worldwide — at $40–150/mo is a comfortable sub-$5M ARR target.
  • Why now for them: The 2026 rule changes made an existing nuisance into an existential one. Portfolio-shared limits mean one careless client number now threatens an agency’s whole Business Manager. The cost of not watching just went up sharply.

5. Product sketch (MVP)

  • Pre-flight template checker: paste a template before submitting to Meta; get a rejection-risk score + the specific reason (category mismatch, format, policy) and a rewritten compliant version.
  • Live quality radar: one dashboard across all your numbers and BSPs (via Cloud API + webhook integrations), showing green/yellow/red per number and per template in real time — not after the email.
  • Drop alerts: instant Slack/WhatsApp/email alert the moment a number’s quality slips to yellow, a template gets paused, or unanswered-message counts spike toward a throttle threshold.
  • Bot-ban drift detector: samples your bot’s outbound replies and flags when conversations look “general-purpose” (open-domain answers) instead of task-specific — the exact behavior Meta’s Jan 2026 ban targets.
  • 48-hour engagement guard: tracks reply rates against the unanswered-message tracking threshold and warns before you cross it.
  • Portfolio health score: a single number per Business Manager so an agency sees its riskiest client at a glance.
  • Recovery playbook: when something goes red, a step-by-step “do this now” checklist (pause the template, clean the list, edit-and-resubmit) instead of guessing.

6. AI angle — what’s load-bearing

Two places AI does real work, not decoration:

  1. Template + reply classification. Predicting whether Meta will reject/pause a template, and whether a live bot conversation reads as “general-purpose” vs “task-specific,” is a nuanced language judgment against Meta’s fuzzy, unpublished criteria. That’s an LLM classification job — rules-based regex can’t catch category drift or open-domain answers. Remove the AI and you’re left with a dumb webhook forwarder anyone could clone.
  2. Compliant rewrite. When a template is risky, generating an approved-style rewrite that keeps the marketing intent is the part agencies will actually pay for — it turns a “you’re at risk” warning into a “here, send this instead” fix.

7. Localization angle

India-first is the natural wedge and it’s a real one. India is the single largest WhatsApp Business API market, the BSP reseller ecosystem (AiSensy, Interakt, Gupshup) is densest there, and pricing tolerance is ₹-scale — a ₹999–3,999/mo tier works where a $49 tier wouldn’t. WhatsApp-native distribution (the buyers live in WhatsApp groups for agencies/resellers) and multilingual template checking (Hindi, Tamil, Telugu, Bahasa, Portuguese for LATAM) are concrete edges over a generic global tool. SEA (Indonesia, Vietnam) and LATAM (Brazil, Mexico) are the obvious follow-on markets — same platform, same rules, same pain, lighter local tooling.

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

  • Pricing: Tiered by numbers watched. Solo/freelancer $39/mo (up to 5 numbers); Agency $99/mo (up to 25 numbers); Studio $199/mo (up to 75 numbers + portfolio score + API). India pricing ₹999 / ₹2,999 / ₹6,999.
  • ACV: ~$1,000–1,400 blended (mix skews to the $99 agency tier).
  • Rough math to $1M ARR: ~850 paying accounts at a blended ~$98/mo. Entirely reachable from the agency/reseller base of the top 5 BSPs.
  • Rough math to $5M ARR: ~4,200 accounts, or fewer accounts + an expansion to per-number usage billing and a “managed recovery” upsell (we file the Meta review / clean the list for you). Add white-label so BSPs themselves resell GreenLight to their own resellers.
  • Expansion path: more numbers per account as agencies grow their book; add list-hygiene/verification, a deliverability benchmark report, and managed-recovery as paid add-ons.

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

  • WhatsApp/BSP reseller communities: AiSensy, Interakt, and Gupshup each run partner/reseller programs and Telegram/WhatsApp groups with thousands of agencies. Get into 10–15 of them, post the free pre-flight template checker, convert the heavy users. These people are the buyer and they congregate.
  • Free template-rejection checker as a lead magnet: a public tool — paste your template, see the rejection risk free. Rank it for “WhatsApp template rejected” / “template paused 132015” searches (high-intent, frequent pain). Gate the multi-number dashboard behind signup.
  • Cold outreach to agencies: scrape the partner directories of the top 6 BSPs and the “WhatsApp marketing agency” listings on Clutch/Sortlist. Send a 60-second Loom: “here’s your client’s number going red — want the alert before Meta does it?” Expect 3–5% reply from people who’ve been burned.
  • Trustpilot/G2 burn list: DM the agencies leaving 1–2 star BSP reviews complaining about template rejections and support delays — they’re pre-qualified pain.

10. Build complexity — justification

Medium. The Cloud API webhook integrations (quality rating, template status, messaging-tier events) are well-documented and off-the-shelf, and the AI classification/rewrite runs on standard LLM APIs. The real work is breadth — supporting multiple BSPs’ auth/webhook quirks, keeping the rejection/ban heuristics current as Meta silently changes the rules, and a clean multi-number dashboard. A 2-person team ships a credible v1 (single-BSP + pre-flight checker + alerts) in ~10–12 weeks; full cross-BSP coverage is the 4–6 month version.

11. Gating checklist

GatePass?Note
Legal in target market✅Reads official Cloud API signals the customer already owns; no scraping, no ToS violation.
Ethical — no harm / dark patterns✅Helps senders comply and respect opt-outs; aligned with Meta’s spam-reduction intent.
Market exists (evidence above)✅Millions of API senders; documented 2026 enforcement pain; loud forum complaints.
1–5 person team can build this✅Off-the-shelf APIs; integration breadth is the only heavy lift.
Launchable with <$50K / ₹40L✅Two builders + LLM/API costs; no capex.

All five pass.

12. Feasibility score

AxisWeightScoreNotes
Problem intensity2016/20Losing a number is hair-on-fire for an agency, but it’s episodic, not daily — they may underinvest until the first ban scares them.
Demand evidence1512/15Strong: enforced 2026 rules, error-code forums, BSP review complaints, reseller communities. Slightly short of a paid direct competitor proving the exact wedge.
Build feasibility1511/15Doable in 12 weeks for single-BSP v1; cross-BSP breadth + drifting Meta heuristics add real engineering.
Distribution clarity1511/15Reseller communities + free checker + burn lists are concrete; conversion math is estimated, not proven.
Revenue mechanics1512/15Clear tiered pricing benchmarked to BSP spend; ~850 accounts to $1M is realistic. Churn risk if agencies treat it as insurance they cancel between scares.
Time to first revenue108/10Free checker → paid in weeks; agencies buy fast, no procurement.
Defensibility105/10Execution + accumulating rejection/ban dataset is a soft moat; a BSP could bolt on a version, but won’t (incentive conflict). Platform-dependency drags it.
Total10075/100

13. Qualitative modifiers

Founder-fit tags

technical-heavy · domain-expertise-required — you need someone who has actually run WhatsApp API campaigns and felt template pauses / quality drops first-hand. Without that domain scar tissue, the heuristics will be wrong.

Key assumptions to validate (3–5)

  1. Assumption: Agencies will pay monthly for prevention, not just churn in after a ban and out once calm. How to test: Pre-sell annual plans to 20 agencies; measure how many commit before any incident.
  2. Assumption: Cloud API webhooks expose quality/template/engagement signals with enough lead time to alert before Meta restricts. How to test: Instrument 10 live numbers for 30 days; measure alert-to-restriction lead time.
  3. Assumption: The free template-rejection checker actually converts to paid dashboard signups. How to test: Ship the checker standalone; measure free→paid conversion over 4 weeks.
  4. Assumption: “General-purpose drift” detection is accurate enough to be trusted, not a false-alarm machine. How to test: Hand-label 500 bot conversations; measure classifier precision/recall.

Risk flags

  1. Platform dependency: Entire product rides on Meta’s Cloud API and its undocumented, shifting enforcement. Meta could change signals, restrict webhook access, or ship its own pre-flight checker. Mitigate by being cross-BSP and fast.
  2. Market timing / “insurance churn”: Pain is episodic. Buyers may cancel between scares. Mitigate by making it daily-useful (campaign-time pre-flight checks), not just an alarm.
  3. Incumbent co-option: A BSP could add a basic version; the defense is independence (cross-BSP) and a dataset advantage, but the moat is thin.

14. Structured verdict

Score:                  75/100
Verdict:                GO
Confidence:             Medium
Best-fit builder:       Technical founder who has personally run WhatsApp API campaigns (felt the template-pause / quality-drop pain), plus light distribution hustle into reseller communities
Time to revenue:        6–10 weeks (free checker → paid dashboard)
Capital to launch:      ₹3–6 lakh / $4–7K
Top 3 assumptions to validate first:
  1. Agencies pay for prevention before a ban — pre-sell 20 annual plans, measure pre-incident commits
  2. Webhooks give enough lead time to alert before restriction — instrument 10 numbers for 30 days
  3. Free template checker converts to paid — ship standalone, measure free→paid over 4 weeks
Kill criteria:
  - Abandon if <10% of 50 agencies in reseller communities show interest after seeing a live "your number is going red" alert
  - Abandon if webhook lead time is consistently <2 hours before restriction (no time to act = no value)
  - Abandon if a BSP ships an equivalent cross-portfolio guard before your v1

15. Next step — 1-week validation sprint

  • Day 1–2: Build the standalone pre-flight template checker (paste template → rejection risk + reason + rewrite). One screen, real Meta rejection-reason heuristics.
  • Day 3–4: Drop it into 8–10 WhatsApp agency/reseller communities and DM 30 agencies leaving 1–2 star BSP reviews about template rejections. Pitch: “free check now, full multi-number quality radar coming — want early access?”
  • Day 5: Decide go / no-go on a falsifiable bar: ≥40 checker uses AND ≥10 agencies join an early-access waitlist with a credit card / ₹-deposit pre-order. Below that, the pain isn’t urgent enough to pay for — revisit or kill.

The result is falsifiable: either agencies put down money/intent for prevention before a ban, or they don’t.

Interested in a detailed proposal?

Get a deep-dive with market research, competitive analysis, and implementation roadmap.

Contact us

info@startupbasket.ai