# P13 — Competitor Analysis

**Updated:** 30 August 2026 · Web-verified 30 Aug 2026. Prices from vendor pages and 2026 round-ups; anything quoted to a client must be re-checked on the vendor's own page that week (**[verify]** discipline). Gaps in the data are reported as gaps — they are findings.

---

## 1. The market has two layers — we compete in one, buy from the other

```
Layer 2 (managed services — OUR competitors): Smith.ai, Goodcall, Rosie, Slang.ai,
        Loman-class vertical bots, and hundreds of agencies reselling wrappers
                    ▲  built on
Layer 1 (infrastructure — OUR suppliers): Retell, Vapi, Synthflow, Bland, + wrapper
        middleware (VoiceAIWrapper, Vapify, Trillet-class)
```

## 2. Layer 1 — platform (supplier) comparison

| Platform | Advertised base | Realistic all-in $/min | Numbers | White-label / agency | Notes for us |
|---|---|---|---|---|---|
| **Retell AI** | $0.07/min, pay-as-you-go, $10 free credit | **$0.10–0.18** (their own ecosystem cites $0.13–0.31 with premium components) | $2/mo, telephony $0.015/min | No native white-label; Enterprise from ~$8k | **Primary (D1):** predictable modelling, free concurrency, native booking tools, QA dashboards |
| **Vapi** | $0.05/min orchestration | **$0.07–0.25** (BYO keys: budget ≈ $0.08–0.10) | Via Twilio/BYO | No native white-label; wrappers exist | **Fallback + cost lever:** cheapest controllable stack, most knobs; native Google Calendar tool |
| **Synthflow** | Usage ≈ $0.12/min effective | ~$0.12–0.15 | Bundled | **Reseller toolkit ≈ $2,000/mo** | No-code, agency-marketed. Toolkit price = 15–20 clients of margin → rejected for pilots (D3) |
| **Bland** | $0.09/min (tiered to $0.11–0.12 + plan fees) | ~$0.10–0.15 + ancillary fees (transfer time, $0.015 min/call outbound attempts) | Bundled | Enterprise-oriented | Fee structure penalizes exactly our usage pattern (many short calls, transfers) |

**Sources:** cloudtalk.io (vapi-ai-pricing, retell-ai-pricing, bland-ai-pricing) · cekura.ai (vapi, retell per-minute) · zeeg.me (vapi, synthflow, bland) · layer3labs.io/guides/retell-ai-pricing · autocalls.ai/article/synthflow-review · voiceaiwrapper.com/compare/synthflow-alternative — all accessed 30 Aug 2026.

**Gap/finding:** none of the four publishes a first-party, all-in "$/min for a typical SMB receptionist" figure — every all-in number above is reconstructed from component pricing by third parties. Our own P1 test calls produce the first number we can actually bank on (task `F-04` measures real $/min on our stack).

## 3. Layer 2 — managed-service (real) competitors

| Competitor | Price (2026) | Model | Weakness we exploit |
|---|---|---|---|
| **Smith.ai** | $95/mo Starter → $270 Basic → $800 Pro; ~$1.60–1.90/call AI, $3 human transfer | AI + human backup, US-centric, strong brand | Per-call pricing is opaque to SMBs; no BD/Gulf presence; no WhatsApp channel; zero hands-on config per client at low tiers |
| **Goodcall** | from $79/mo (caps by unique callers) | Self-serve flat plans | DIY setup burden on the owner; limited vertical depth; no local-market service |
| **Rosie** | from $49/mo | Budget self-serve | Same — template answers, no integrations depth, no managed QA |
| **Slang.ai** | $399–599/mo/location (Core/Premium; ~$379 annual) | Restaurants only, managed-ish | Price excludes small independents — our restaurant tier is 1/3 the cost |
| **Loman/Hostie-class vertical bots** | ~$100–300/mo | Vertical SaaS (restaurants/dental) | US-only, single-vertical, no local relationship |
| **Agencies reselling Synthflow/Vapi wrappers** | Wildly variable, commonly $200–500/mo + setup $500–2,500 | White-label wrapper + markup | Most are prompt-and-pray operations with no QA discipline and high prices to cover the $2k/mo toolkit; their existence proves demand |
| **In BD/Gulf specifically** | **Gap:** no established Bangla/Bangladesh AI-receptionist service found in research; Gulf has early Arabic offerings but no dominant SMB brand | | First-mover room, but also a caution: absence may mean telephony friction + willingness-to-pay is untested — exactly what the pilot measures |

**Sources:** withallo.com/blog/ai-receptionist · agentzap.ai (pricing guide, $25–899 landscape) · workflowstackai.com/blog/ai-receptionist-comparison-2026 · contractortoolstack.com/software/smith-ai · bitebuddy.ai/blog/slang-ai-pricing · loman.ai · trillet.ai (white-label landscape) — accessed 30 Aug 2026.

## 4. Where a Dhaka-cost agency wins (and where it doesn't)

**We win when:**

| Edge | Why it's real |
|---|---|
| Managed care at self-serve prices | Weekly human QA + prompt tuning costs us $3–8/h of skilled time; a US competitor pays 5–10×. At $99–149 we deliver what costs them $300+ to deliver |
| Warm local relationships | The first 3 clients already trust ThemeTags with their restaurant system/CRM/marketing. Smith.ai cannot call them |
| WhatsApp-first design | BD/Gulf SMBs live on WhatsApp; US incumbents are phone-line-first. WhatsApp Business Calling (inbound free) sidesteps the whole UAE/BD number problem |
| Integration hands | We already build their CRMs — hooking the lead webhook into a system we wrote is an afternoon |
| Bundling | Receptionist + our CRM + marketing retainer = a stickier composite no US point-solution matches |

**We do NOT win on:** brand trust with cold US buyers (Smith.ai owns that), human-backup hybrid (we have no call-center bench — never promise human answering), enterprise/multi-location, or price against $49 self-serve (we don't chase buyers who won't pay for managed care).

## 5. Positioning sentence

*"The managed AI receptionist for local businesses — set up for you, tuned every week, on WhatsApp and phone, from the agency that already runs your systems — at a third of the US price."*