1. Verdict
Why a ladder instead of one target
This idea is assessed at six MRR levels instead of one fixed target. A consumer subscription product's outcomes are highly skewed: most products reach little, some reach a comfortable monthly figure, and very few break out. A single "$20k MRR" number hides the most useful information: how likely each level is, and what each level costs in time.
The anchor: what happens to new subscription apps
RevenueCat's State of Subscription Apps 2026 (all categories, apps on its platform):
- 17.3% of newly launched apps reach $1k monthly revenue within 2 years (19% the year before).
- 4.6% reach $10k within 2 years (5.3% the year before).
- AI-powered apps earn 41% more per payer but churn about 30% faster.
Education-category figures from the same report:
- Download-to-trial: 6.5% median.
- Revenue per install at day 14: $0.30.
- Realised LTV per payer after 1 year: $22.82.
- Median annual price: $39.99–44.99.
- Annual-plan renewal: 24% median, the weakest of any category.
Three things push the odds down from the base rate:
- The category leader (Duolingo) gives the core mechanic away free.
- Indian prices are ₹199–299 a month.
- Education renewal is weak.
Three things push them up:
- Spoken English is a proven, paid need in India.
- The recommended version serves a group with a deadline (job seekers).
- Institutes can buy seats in bulk.
Odds at each MRR level (reached by month 24)
| MRR at month 24 | Base rate: new subscription apps | As described (3 languages × 2 goals, game city, everyone) | General spoken English (Hindi-medium, head-on with SpeakX / Stimuler / Duolingo) | Heritage Hindi or Bengali for diaspora children (USD, parents pay) | B2B only (institutes and employers) | Recommended, ~15 hrs/week | Recommended, full-time |
|---|---|---|---|---|---|---|---|
| $1k | 17.3% | 15% | 25% | 28% | 25% | 35% | 50% |
| $3k | — | 7% | 13% | 14% | 17% | 22% | 33% |
| $5k | — | 4% | 8% | 8% | 11% | 15% | 24% |
| $10k | 4.6% | 2% | 4% | 4% | 6% | 7% | 13% |
| $20k | — | ~1% | 2% | 1.5% | 3% | 3% | 6% |
| $50k | — | <0.5% | <1% | <0.5% | 1% | 1% | 2% |
Recommended means work-focused spoken English for Hindi- and Bengali-medium learners, as set out in WHAT-TO-DO-AND-AVOID.md, with the 3 conditions met:
- a spoken-English trainer as advisor;
- a sustained content routine;
- 50 paying before building.
If any one condition is missing, multiply the recommended odds by about 0.65.
Notes on the alternatives:
- General spoken English faces SpeakX (about 200k payers), Stimuler, ELSA and a free Duolingo, and the category buys growth with heavy ad spend.
- Heritage learning earns about 3–4x more per payer in USD, and Duolingo has no Bengali course. But the audience is small, the users are children (COPPA and DPDP parental consent), children's Bengali speech recognition is weak, and no funded company has shown demand.
- B2B only is the most reliable per rupee earned. It needs regular sales meetings and long procurement cycles, which scale with hours worked. That is why it is added to the recommended version from month 9 rather than leading.
These are judgment estimates anchored to the RevenueCat base rate, not computed statistics.
How to read the ladder
- $1k–3k is a realistic outcome if learners can see their speaking improve. It takes about 400–1,200 payers, which organic content and shares can deliver.
- $5k–10k needs institute seats on top of individual payers. Consumer volume alone at ₹199 means 2,000–4,000 payers with ~12% monthly churn.
- $20k and above needs paid acquisition that pays back. Seekho spent ₹134 Cr on ads for ₹142 Cr of revenue in FY25. That's a different operating mode.
- The odds fall fastest between $3k and $5k. That is where organic channels usually stop being enough at Indian price points.
Portfolio definitions (for comparison with other ideas)
| Tier | Definition | Recommended, ~15 hrs/week | Recommended, full-time | As described |
|---|---|---|---|---|
| Tier A (venture) | $1M ARR within 36 months | <1% | ~1% | <0.5% |
| Tier B (profitable business) | $20k+ MRR within 24 months, profitable, ≥80% gross margin | ~2% | ~4% | <1% |
Tier B is lower than the $20k rung in the ladder. The ladder counts gross margin ≥70%. Voice and model costs at Indian prices make ≥80% hard to reach until volume allows self-hosted open models, which is what Duolingo did to get under $0.01 a call.
How this compares with the other ideas in this portfolio: lower than CreatorDesk and TripDesk (B2B), and lower than AstroCompanion at every level. That category already pays per question at ₹210 an order, and it has no free incumbent doing the same job.
What this idea offers instead:
- a clearly proven need (spoken English for work);
- first revenue within weeks through a WhatsApp concierge;
- a B2B seat line that doesn't depend on ads.
Fundable in current form?
No. Duolingo (58.7M DAU, $1.2B revenue guidance for 2026) is moving AI conversation into cheaper tiers. Speak is valued at $1B, Loora raised $22M in September 2026, and SpeakX raised $16M to cover exactly this audience in regional languages. Nothing in the idea as described is defensible yet. See 06-fundability.md. The business doesn't need funding to reach the levels where its odds are reasonable.
One-paragraph verdict
The problem you describe is real, but it's a slightly different problem from the one you wrote down. Leading apps do place learners above zero, but they place them by reading. The learner who understands English and still freezes in an interview, on a customer call or in a meeting is poorly served, especially if they think in Bengali or Hindi.
The city map and voice companion are good delivery, but Duolingo already ships scene-based speaking for free. They become a business only when the places on the map are the learner's real, high-stakes situations and the app proves the learner is improving.
Build that for Hindi- and Bengali-medium learners heading into work. Put the paid core on the web and WhatsApp, and add institute seat licences from month 9. On those terms, $1k–3k MRR is a reasonable expectation, $5k–10k needs the institute line, and $20k+ is a stretch that needs paid growth.
Teaching Hindi and Bengali as new languages, and the casual-conversation track, should wait until learners ask for them.