How We Test Spanish Learning Apps
We are not sponsored by any app. We earn commission on Babbel, Pimsleur, Rosetta Stone, Rocket Spanish, Lingopie, italki, and Preply — not on Duolingo. Our scoring is independent of our commission structure.
Testing Protocol
Minimum test period: 90 days per app
Every app in our reviews was used as a primary study tool for a minimum of 90 days before scoring. We did not supplement with other apps during the primary test period to isolate each app's effectiveness.
The 50-sentence speaking test
We use a standardised set of 50 sentences across all app reviews. The sentences cover ordering food, asking directions, introducing oneself, handling a misunderstanding, and discussing plans. We attempt to produce each sentence from memory after 45 hours of study with each app. Results are reported as sentences produced correctly / 50.
Scoring axes (6 total, equal weight)
- Speaking output quality — can the learner produce phrases under light pressure after 45 hours?
- Grammar scaffolding — does the app teach structure, or just patterns?
- Pronunciation feedback — is speech recognition meaningful, or binary pass/fail?
- Content ceiling — how far can the app take a learner before hitting its limit?
- Value relative to price — at the real discounted price, is this better or worse than alternatives at the same price point?
- Realism of marketing claims — does the app's marketing match what it actually delivers?
Disclosure
We earn affiliate commission on qualifying purchases via links to: Babbel (CJ affiliate, ~$15-25/sale), Pimsleur (CJ affiliate), Rosetta Stone (CJ affiliate), Rocket Spanish, Lingopie (PartnerStack, 30% MRR 12 months), italki (ShareASale, 10% first lesson), Preply (Awin, 10% first purchase).
We do not earn commission on Duolingo (their affiliate program is not available to independent review sites at our scale), Mango Languages, Anki, or Dreaming Spanish. We recommend all of these when they are the right answer for a given user situation, including when that answer is "don't spend any money."
Our overall scores and narrative recommendations are written independently of our commission structure. The Duolingo review score (6/10) is lower than the Pimsleur score (8.5/10) not because Duolingo doesn't pay us, but because Pimsleur produces better speaking outcomes per hour of study. We document our evidence for this claim in each review.
Update Log
- 2026-05-20 — Site launch. All 5 initial reviews and 2 comparisons published.
- 2026-04-15 — Pimsleur review finalised (90-day test complete).
- 2026-03-15 — Babbel review finalised.
- 2026-02-28 — Rosetta Stone review finalised.