GTM Playbook
13 verticals · Deal tracker · ROI calculator · Email templates · Pricing · Demo scripts · Partners · Expansion
v4 includes: Follow-up email templates, pricing & commercials, live demo scripts, win/loss tracker, call scorecard, SI partner list, co-sell motion, 90-day onboarding, expansion playbook, and regulatory calendar.
113
Languages
50ms
ASR Latency
13
Verticals
Live
api.valsea.app
Priority Matrix
| Vertical | Priority | Market | Pipeline | Top Use Case | Markets |
|---|
ROI Calculator
13 vertical-specific models — select your prospect's industry
Pricing & Commercials
Internal guidance v5.0 — March 2026 · DO NOT share externally
🔒 Internal use only. This page reflects the V5.0 pricing strategy. The rate card must NOT go live for Japanese, Korean, or Singlish routes until COGS are confirmed.
Subscription Tiers
Free
$0
55 min/month · hard cap
- No batch, no diarisation, no streaming, no SLA
- Usage nudge at 44 min (80%)
- COGS: ~$0.14/user/month
Starter
$99/mo
5,000 min included · eff. $0.0198/min
- Overage: $0.022/min RT
- Developers, small teams
- ⚠️ Review: $0.022 overage may drive churn
Growth ★
$299/mo
20,000 min · eff. $0.0150/min
- Overage: $0.018/min RT
- Agencies, meeting intelligence
- 121 customers to cover $30K OpEx
Business
$1,500/mo
120,000 min · eff. $0.0125/min*
- Overage: $0.015/min RT
- Contact centres, enterprise
- 99.9% SLA included
Enterprise
Custom ≥$5K/mo
12-month minimum commitment
- Private deployment · custom SLA
- Annual prepayment: 15% discount (CFO sign-off)
- Annual escalation capped at CPI+3%
COGS Gate
The $0.013/min rate card must NOT go live for the following routes until conditions are met.
| Provider / Route | COGS/min | Gross Margin | Status | Condition |
|---|---|---|---|---|
| Qwen3 — Chinese/Bahasa | $0.0020 | 84.6% | ✓ Live | Default routing. |
| Blaze — Vietnamese | $0.0050 | 61.5% | ✓ Live | Non-negotiable VI routing. |
| AmiVoice — Japanese (RT) | $0.0165 | −26.9% ⚠️ | 🔴 Blocked | Negotiate to ≤$0.010 or apply surcharge. |
| Naver CLOVA — Korean | TBD | Unknown | 🔴 Blocked | Commercial agreement not signed. |
| MERaLiON — Singlish/Tamil | TBD | Unknown | 🔴 Blocked | API trial only. No commercial agreement. |
Competitor Pricing
| Provider | Base Rate/min | All-in Production Rate | vs VALSEA |
|---|---|---|---|
| VALSEA | $0.013 RT | $0.022 (RT+diar+stream) | — |
| Google Chirp | $0.016/min | $0.016/min | 27% cheaper |
| Azure Speech | $0.0167/min | $0.0167/min | 24% cheaper |
| Speechmatics | $0.0117/min | $0.0117/min | VALSEA 11% pricier (no SEA) |
| Amazon Transcribe | $0.024/min | $0.024/min | 46% cheaper |
Pilot-to-Paid Conversion Ask
Always Propose This
"The pilot is free — 2 weeks, your audio data, one API key. At the end we have a WER comparison and a cost model. If the numbers work, we convert to Growth ($299/mo). No commitment until you've seen the data."
Internally: Pilots that hit the WER improvement threshold convert at ~70%. Push hard for the pilot — it IS the sales motion.
Follow-Up Email Templates
Select vertical → select template type → copy and personalise
Vertical
Live Demo Scripts
Step-by-step scripts for running a live VALSEA demo against competitors
Golden rule: Always demo on the prospect's own audio — not a VALSEA sample. Ask them to bring 10 minutes of real call recordings.
The Side-by-Side WER Demo
Step 1
Before the meeting — prepare your audio
Email the prospect 48hrs before: "Could you bring 10 minutes of real call recordings in Vietnamese/Thai? We'll run a live accuracy comparison against your current ASR." If they can't, have 3 pre-approved samples ready.
Step 2
Open the comparison frame
Pull up two browser tabs: tab 1 = prospect's current ASR provider. Tab 2 = api.valsea.app. Have both transcription inputs ready before they arrive.
Step 3
Run the audio — let silence work
Play the audio. Both transcripts appear in real time. Don't narrate. Let them read for 20 seconds. Financial terms and Vietnamese/Thai words will be visibly more accurate.
"Notice [specific term they use] — it's transcribed as [wrong] here and [correct] here. In a compliance context, that's the difference between a clean audit and a breach."
Step 4
Quantify the gap
Pull up the ROI calculator (already filled in for their vertical). Show the cost model based on their call volume.
"At your call volume, this accuracy difference costs you approximately $[ROI number] per year. The pilot removes all doubt."
Step 5
Close the pilot
Slide the pilot success criteria in front of them. Ask them to agree on what success looks like — not a contract, just alignment.
"Can we agree: if VALSEA hits [WER target] on your audio by end of week 2, you'll bring this to your CTO/CFO for a contract discussion?"
Competitor-Specific Demo Angles
| If they use… | Demo angle | Line to use |
|---|---|---|
| Google STT | Run same Vietnamese financial call through Google demo + VALSEA. Count mis-transcribed financial terms. | "Google is world-class for English. For Vietnamese financial vocabulary, count the errors yourself — that's your compliance exposure." |
| AWS Transcribe | Show AWS has no Vietnamese language model — it defaults to English. | "AWS Transcribe literally doesn't support Vietnamese. Whatever your team is doing today, it's guesswork." |
| Whisper (OpenAI) | Show Whisper batch-only limitation. Run VALSEA real-time while Whisper processes a file. | "Whisper is a great research model. It has no real-time capability — you can't build a live agent assist on it." |
| In-house/custom | Ask what the WER is on Vietnamese today. They usually don't know. Offer to measure it during the pilot. | "Let's just measure it. If your WER is better than ours, keep it. If not, you'll know in 2 weeks." |
Win / Loss Tracker
Log every deal outcome — patterns in objections become the most valuable part of this playbook
Log every deal. Over time, the pattern of what kills deals — and what wins them — is more valuable than any script.
Log a Deal Outcome
Call Scorecard
Complete this immediately after every prospect call — honestly, not aspirationally
If you score below 60%, review the discovery question bank and pitch guide before the next call with this account.
0 / 15
Complete the checklist above
Partners & Co-sell Motion
SI partners by SEA market, cloud co-sell angles, and telco channel positioning
System Integrator Partners
Cloud Co-sell Positioning
| Cloud Partner | Co-sell angle | How to position VALSEA | Who to engage |
|---|---|---|---|
| AWS | VALSEA is AWS-native. AWS Marketplace candidate. | "VALSEA runs on EKS, stores in S3, ships to ap-southeast-1. Your AWS team gets a workload win." | AWS ASEAN FSI / Telco PDM |
| Azure | VALSEA as Nuance replacement for SEA languages. | "Nuance has no production Vietnamese or Thai model. VALSEA fills the gap — same Azure infrastructure." | Microsoft ASEAN AI Partner team |
| Google Cloud | VALSEA as complement to CCAI for SEA. Replaces the ASR layer. | "VALSEA plugs into CCAI as the ASR layer for Vietnamese and Thai." | Google ASEAN CCAI partner team |
Telco Channel Play
Key insight: SEA telcos want to package AI services for enterprise customers. VALSEA can be white-labelled as a "Voice AI API" in their enterprise product catalogue. Sign 1 telco, reach their 500 enterprise customers.
Post-Sale & Expansion Playbook
90-day onboarding checklist and land-and-expand motion from 1 use case to 3+
90-Day Onboarding Checklist
Land-and-Expand Motion
Every VALSEA deal should expand from the landing use case to at least 3 use cases within 12 months. This is where LTV is built.
Regulatory Calendar
Compliance deadlines that create buying urgency — reference in discovery and follow-up emails
How to use: Before any FSI, Telco, or Government meeting, check if there's an upcoming deadline in their market. A compliance deadline is the best sales accelerant — it creates urgency that isn't manufactured.
Universal Pitch Guide
How to open, pitch, and close any vertical — in 3 moves
Golden Rule: Never lead with technology. Always lead with the metric the buyer cares about — cost, compliance, speed, or revenue.
Move 1
Name the Pain
- Cite a specific, quantified problem they recognise
- Reference their language, market, customers
- Make them say "yes, that's us" in 30 seconds
Move 2
Show the Gap
- Explain why Google/AWS/Whisper fail on this
- Name the competitor and the WER gap
- Show accuracy on their actual language
Move 3
Close With Proof
- Offer a demo on their own audio
- Propose a 2-week pilot with a clear metric
- Trial access available same day
Standard Pilot Ask
"Let us run a 2-week pilot on your real audio data. We'll benchmark VALSEA accuracy against your current solution on Vietnamese / Thai / Singlish. You'll have a Word Error Rate comparison and a cost model by end of week 2. Zero commitment, zero infrastructure change — one API key."
Objections & Answers
Every objection VALSEA will face — and how to handle it
"We already use Google STT / AWS Transcribe / Whisper."
Those tools are excellent for English. On Vietnamese, Thai, and Singlish their WER is 25–40% higher than VALSEA's specialist models. Ask them to test on 10 minutes of their own audio and compare side by side.
"You're a startup. What's the reliability guarantee?"
VALSEA runs on production Kubernetes on AWS with a multi-account architecture. We offer SLA commitments and direct technical support. Happy to arrange a technical call before any commitment.
"Our voice data can't leave the country."
VALSEA supports data residency configurations — audio processed within approved geographic boundaries, discarded without retention. We align with MAS, BOT, SBV, and OJK. This is solvable, not a blocker.
"We don't have resources to integrate this."
VALSEA is OpenAI-compatible — if your team has used any OpenAI API, the pattern is identical. Most teams are live in under a day.
"The ROI isn't clear enough to justify the switch."
Propose the 2-week pilot. At the end you'll have real accuracy data on their audio, a cost model vs. their current solution, and a projected ROI based on actual call volume. The pilot is the business case.
"Speechmatics is cheaper."
Speechmatics charges $0.0117/min but covers no SEA languages. If you need Vietnamese, Bahasa, Singlish, Korean, or Japanese, there is no comparable product at any price.
"AssemblyAI starts at $0.002/min."
That is the async base rate without diarisation or enhanced models. In production, AssemblyAI reaches $0.018–0.025/min all-in. VALSEA all-in is $0.022/min — competitive, with measurably better accuracy on your target languages.
"Human transcription is good enough."
Human transcription costs $3–4/min. VALSEA batch is $0.010/min — 97% cheaper. For 1,000 hours/month, the annual saving exceeds $2.1M.