El intérprete de pruebas A/B: Lanzar, Mantener o Descartar en un solo Prompt
Why this prompt matters
Bad ship/kill calls are expensive in both directions. Shipping a false positive locks in code that looks like a winner in the dashboard but quietly erodes revenue for months before anyone traces it back to the test. Killing a real winner because the raw numbers looked shaky under-delivers growth you already earned. And teams that 'peek' at results early and stop tests as soon as the trend looks good introduce a well-documented bias that inflates the apparent win rate of every experiment program — this prompt forces a power and significance check before any of that can happen.
What we use it for
You're a growth PM and your two-week checkout redesign A/B test just wrapped. Leadership wants a ship/no-ship call in tomorrow's roadmap review, but the raw dashboard just shows a conversion rate that went up — it doesn't tell you if that's a real 5% lift or noise that will vanish next month, and nobody on the team has time to run the statistics by hand before the meeting.
Prompt
Act as a senior product analyst and growth data scientist with deep expertise in experimental design and statistical inference for digital products. CONTEXT: I ran an A/B test with the following setup: - Test name / hypothesis: [TEST NAME/HYPOTHESIS] - Primary success metric: [PRIMARY SUCCESS METRIC] - Control group result: [CONTROL METRIC AND VALUE] - Variant group result: [VARIANT METRIC AND VALUE] - Sample sizes: [SAMPLE SIZES] (control vs. variant) - Test duration: [TEST DURATION] - Guardrail metrics tracked (if any): [GUARDRAIL METRICS, IF ANY] - Any other context (traffic source, platform, seasonality, known anomalies): [ADDITIONAL CONTEXT] TASK: Analyze this experiment and produce a ship/hold/kill recommendation. Your analysis must: 1. Calculate or estimate statistical significance (p-value or confidence interval) for the primary metric given the sample sizes provided. 2. Assess statistical power — flag explicitly if the sample size is too small to trust the result either way. 3. Distinguish statistical significance from practical/business significance — a "significant" result that moves the needle by a trivial amount is not a reason to ship. 4. Check the guardrail metrics for regressions, even if the primary metric looks like a win. 5. Give a clear final recommendation: SHIP, HOLD (extend test), or KILL — with your reasoning stated in plain terms a non-technical stakeholder can follow. CONSTRAINTS: - Do NOT recommend shipping on a result that is underpowered — say so explicitly and recommend the minimum additional sample size or duration needed instead. - Explicitly flag possible novelty effects if the test ran less than 2 full business cycles (e.g., under 2 weeks for a weekly-cadence product). - Explicitly flag risk of "peeking" bias if the numbers suggest the test may have been stopped early based on interim results. - Never conflate correlation from a single test with certainty — mention confounding variables if the context suggests any (seasonality, traffic mix changes, etc.). - Keep the tone direct and decision-oriented — this is going into a roadmap review, not a stats textbook. OUTPUT FORMAT: ## Statistical Read [Confidence interval / p-value, power assessment, in plain language] ## Business Interpretation [What this actually means for the product and the metric that matters] ## Recommendation: [SHIP / HOLD / KILL] [2-3 sentence justification] ## Caveats & Follow-Up [1-2 risks or suggested follow-up tests]
Result
Scenario: Checkout flow redesign test, e-commerce site, 14-day run.
Statistical Read
Control: 4.8% conversion (n=24,600 sessions). Variant: 5.05% conversion (n=24,450 sessions). That's a +5.2% relative lift. Running a two-proportion z-test on these numbers gives a p-value of approximately 0.041 — just inside the conventional 0.05 significance threshold, with a 95% confidence interval on the lift of roughly [+0.1%, +10.3%] relative. Post-hoc power at this effect size and sample size lands around 52%, well below the 80% threshold you'd want for a decision this consequential. In plain terms: the result is nominally significant, but the confidence interval is wide enough that the "true" lift could be barely above zero or as high as 10% — you don't have enough data to know which.
Business Interpretation
A +5.2% relative lift on checkout conversion is commercially meaningful if real — at your traffic volume that's roughly 60 incremental orders per day. But the wide confidence interval means you're not choosing between "ship" and "don't ship," you're choosing between "possibly a strong winner" and "possibly noise." The guardrail metric (average order value) held flat at -0.3% (not significant), so there's no evidence the variant is cannibalizing basket size to win on conversion rate. No red flags there.
Recommendation: HOLD
Extend the test by at least 7-10 more days to reach ~80% power at the observed effect size (approximately 34,000 sessions per arm). The 14-day window also only covers two weekly cycles, borderline for ruling out a novelty effect on a redesigned flow — extending the window addresses both the power problem and the novelty-effect risk simultaneously. Shipping now on a 52%-powered result risks locking in a false positive that erodes checkout conversion once the novelty wears off.
Caveats & Follow-Up
Check whether the lift is concentrated in mobile vs. desktop traffic — redesigns often show device-specific effects that average out misleadingly. Also confirm the test wasn't peeked at and stopped early; if the 14-day window was chosen after seeing early results trending positive, treat this read with additional skepticism and weight the extension recommendation even more heavily.
La mayoría de los equipos leen un resultado de prueba A/B de la misma manera: número de control, número de variante, el que sea más alto gana. Así es como se lanzan falsos positivos y cómo los verdaderos ganadores son descartados por partes interesadas nerviosas que ven una muestra pequeña y se asustan. Este Prompt existe para forzar cinco comprobaciones que un vistazo al dashboard omite — significancia, potencia, significancia práctica, regresiones en indicadores de protección y riesgos de sesgo como efectos de novedad y visualización anticipada — antes de que alguien se comprometa a una decisión de lanzar, mantener o descartar.
Por qué funciona la estructura
La sección de Contexto pide deliberadamente los tamaños de muestra y la duración de la prueba desde el principio, no solo las métricas principales. Los cálculos de significancia y potencia no tienen sentido sin ellos, y es exactamente la información que la mayoría de la gente omite cuando pegan "el control fue del 4.8%, la variante del 5.05%" en una ventana de chat esperando un veredicto.
La sección de Restricciones es donde reside el valor real. Prohibir explícitamente una recomendación de "lanzar" en un resultado con potencia insuficiente detiene el modo de fallo más común: una prueba que alcanzó significancia nominal (p justo por debajo de 0.05) pero que nunca se ejecutó lo suficiente para alcanzar un tamaño de muestra fiable. La bandera de efecto de novedad señala rediseños y cambios de interfaz que elevan una métrica temporalmente porque los usuarios reaccionan a algo nuevo, no porque el cambio sea realmente mejor — una distinción que solo se manifiesta si la prueba dura lo suficiente para cubrir múltiples ciclos completos de uso. La bandera de sesgo de visualización anticipada aborda una trampa estadística bien conocida: los equipos que revisan los resultados a diario y detienen la prueba en cuanto se ve bien están ejecutando docenas de pruebas de hipótesis ocultas, lo que infla la tasa de falsos positivos muy por encima del 5% nominal.
La sección de Formato de salida separa intencionadamente la "Lectura estadística" de la "Interpretación empresarial". Un valor p de 0.03 no significa nada para un vicepresidente en una revisión de hoja de ruta, pero "hay una posibilidad real de que este incremento sea cercano a cero" sí significa algo. Separar ambos obliga al modelo a traducir las estadísticas en una frase utilizable para la decisión en lugar de detenerse en el número.
Cómo usarlo bien
Rellena los campos entre corchetes directamente desde la exportación de tu plataforma de experimentación — no redondees los números manualmente, ya que el cálculo de la potencia depende de los tamaños de muestra reales. Si no tienes una métrica de protección formal definida, escribe "ninguna registrada" en lugar de dejarlo en blanco — esa laguna merece ser señalada en la salida, ya que lanzar basándose en una sola métrica sin controles de riesgo secundarios es un riesgo en sí mismo.