B Babelio · Operating Playbook 05 · Pricing
refresh: quarterly
Artifact 05 — Demand Engine

Pricing & Packaging

Three tiers, decided. Anchored on one ARPU of $12/mo, gated on the value metric that maps 1:1 to cost — translated audio-minutes, split into cheap subtitle minutes (the monetized core for immersion learners) and metered dub minutes (priced to protect margin).

Pro ARPU
$12/mo
Gross margin (Pro)
73%
CAC payback
4.0mo
LTV : CAC
~3:1
Bar for good A new salesperson can quote any deal — tier, annual discount, seat band, dub overage — from this page on Monday without checking with the founder. Every price is decided, not proposed.
Purpose. Tiers decided with rationale, so packaging is a constraint the team executes against — not a debate it reopens.

1 · The three tiers

Goldilocks structure — the middle tier (Pro) is the plan of record. Free proves the magic and seeds virality; Team is a clearly-secondary margin hedge for creators and small remote teams, not the plan we sell by default.

Free Pro PLAN OF RECORD Team / Prosumer
Price $0 $12/mo$120/yr · −17% $24/seat/moannual only · "Contact sales" >5 seats
Value-metric gate
  • 60 subtitle-min/mo
  • 15 dub-min/mo
  • 2 languages · 1 app at a time
  • ~2,000 subtitle-min/mo
  • 300 dub-min/mo, then $0.06/min
  • all languages · all apps
  • pooled minutes across seats
  • higher dub allotment
  • admin + voice-clone (roadmap)
Features Core subtitle & dub modes. No export. Dual-track overlay + Anki / export. Everything Free has, uncapped languages/apps. Everything in Pro + seat admin, pooled billing.
Target user tire-kicker, casual student Immersion learner (primary ICP) creators / clippers, small remote teams (hedge)
Anti-pattern check — passed The Free tier is constrained on the value metric that maps to COGS (subtitle-min + dub-min + export), not on cosmetic features. The 15 free dub-minutes prove the magic; the meter forces upgrade exactly where cost spikes. Subtitle mode is the monetized core for immersion learners — generous-but-metered for acquisition; dub is metered to protect margin.

2 · COGS math per tier

Single source of truth for AI cost = research/tech.md §8 (Cartesia Sonic $0.006/min + Deepgram Nova-3 + Gemini Flash-Lite, 55% speech duty cycle). The earlier $12.5/hr retail-ElevenLabs framing is deleted — Cartesia retail is margin-viable, so self-host TTS is a Growth-stage option, not a survival requirement.

Per active-hour unit cost (the base rate) Dub ≈ $0.50 / active-hr (STT $0.55 + MT ~$0 + Cartesia TTS $0.36, ×55% duty). Subtitle ≈ $0.31 / active-hr (STT-only, ×55%) — the cheap-tier escape hatch.
Tier · typical usage AI cost Infra + support Total COGS Price GM%
Free60 sub + 15 dub min $0.43 $0.10 $0.53 $0.00 CAC
Pro6 sub-hr + 2 dub-hr (typical) $2.86 $0.40 $3.26 $12.00 73%
Proheavy dub (20 dub-hr) ⚠ $10.00 $0.50 $10.50 $12.00 13%
Teamper seat, blended $5.20 $0.80 $6.00 $24.00 75%
Blended (large free subtitle base) $12.00 65–70%
The one row that breaks the model A heavy-dub Pro user (20 dub-active-hrs ≈ $10 COGS on a $12 plan) is near-unprofitable — which is exactly why dub is metered with $0.06/min overage and free dub is capped at 15 min. Subtitle-led usage keeps GM ≥70%. The binding constraint is dub-minute mix, not TTS unit cost.

3 · Discount ladder

A salesperson can grant anything above the gate line without asking. Below the line needs founder approval. Annual is the only contract length sold for Team.

Standing discounts — no approval needed
1-year annual prepay −10% any rep
2-year prepay −20% any rep
100+ seats (Team) −30% any rep
Anything beyond −30%, stacked discounts, or free overage waivers founder approval escalate

4 · Expansion vector

Pick ONE default expansion vector. The choice is forced by the ICP: immersion learners are individuals, so seats don't expand; the value (and the cost) scales with how much they translate.

Default expansion → USAGE (metered dub-minutes)

Net revenue grows as a learner shifts from subtitle to dub.

A Pro user who exhausts 300 dub-minutes pays $0.06/min overage — the one mechanism that turns a heavier, higher-cost user into a higher-revenue one, keeping NDR climbing from 95% toward 105% without re-pricing the base. It maps 1:1 to the COGS driver, so margin is preserved as revenue expands.

Why not seats: the primary ICP is a solo learner — seats only exist on the secondary Team hedge, so seat-expansion can't be the engine. Why not modules: voice-clone / tone-register are roadmap up-sells, too thin and too far out to anchor expansion today.

5 · Van Westendorp — the #1 validation gate

The price is anchored before the WTP test is run. The single most important pre-pricing experiment is to confirm enough buyers clear the gap between the consumer WTP ceiling and the dub-COGS floor. Run this script on 30–50 immersion learners this week.

"Imagine Babelio subtitles or dubs any app on your PC in real time, in your language, so you can learn from native media."

  1. At what monthly price would Babelio start to feel too expensive to consider?
  2. At what price would it feel like a bargain — clearly worth it?
  3. At what price would it feel so cheap you'd doubt the quality or latency?
  4. At what price is it expensive but still worth considering?

Plot four cumulative curves → "too cheap" × "too expensive" = lower bound; "bargain" × "expensive" = upper bound. Hypothesis: the acceptable band sits around $9–19/mo, with the Optimal Price Point (OPP) near $9–12 and a sharp "too expensive" wall above ~$15 for free-extension-anchored consumers. $12 sits inside the band, just below the wall — deliberately.

Instrument (b) — metered $5/hr concierge test, Week 2 Offer 30–50 buyers a manually-fulfilled metered plan and specifically measure metered-dub WTP vs the subtitle-only ceiling. Apply the <30% kill-criterion: if dub-minute attach or usage can't sustain margin within the WTP band, dubbing stays a subtitle-funded loss-leader and the model becomes subtitle-metered + export only.

6 · Why $12 for the middle tier

Anchor logic · competitor anchor · value-metric scale

$12 is the highest number that sits cleanly inside the validated WTP band ($9–19) yet below the "too expensive" wall (~$15) that free browser extensions like DubTab and Whisperr build in consumers' heads. It anchors against the nearest paid competitors — Transync at $8.99 and Maestra Lite at $23 — landing Babelio above the throwaway-cheap tier (signalling real quality and sub-700ms latency) but well under the prosumer ceiling. Critically, $12 is the price at which the value-metric math closes: a subtitle-led immersion learner (6 subtitle-hrs + 2 dub-hrs/mo) costs ~$3.26, leaving a 73% gross margin and 4.0-month CAC payback. Going to $15 would chase a few extra dollars of ARPU straight into the consumer wall and depress conversion; going to $9 would forfeit margin we don't need to. $12 is the value-capture maximum that still converts.

Pricing don'ts

Don't flat-rate dub minutes

A flat plan on usage-heavy AI is margin death (the Replit lesson). Dub stays metered with $0.06/min overage, always.

Don't lock tiers before the WTP test

$12 is anchored, not final. The Van Westendorp + concierge run on 30–50 buyers is the gate — apply the <30% kill-criterion before scaling dub.

Don't lead with the Team tier

Team is a secondary margin hedge for creators and small teams — not the plan of record. The default sale is Pro at $12.

Linked artifacts
  • 10 · Financial ModelARPU $12, GM 73%, COGS $3.26 and CAC $35 here must match the financial model's assumption cells.
  • 06 · GTM MotionEvery quote in the sales scripts uses the tiers and discount ladder on this page.
  • 01 · ICP BriefThe immersion-learner ICP that gates the value metric is defined here in full.
Babelio · Operating Playbook · 05 — Pricing & Packaging Refresh quarterly · source of truth: research/monetization.md + research/tech.md §8