<p>Transforme qualquer Data Dump em um Boardroom-Ready Insight Report em minutos</p>
Why this prompt matters
Most teams have more data than they have time to analyze. The average analyst spends 40-60% of their time cleaning and summarizing data, and only 10-20% on actual interpretation. Without this prompt, that export sits in a folder until someone has bandwidth — which usually means the insight arrives after the decision was already made. This prompt collapses the gap between raw data and actionable understanding from hours to minutes.
What we use it for
You're a product manager going into a quarterly business review with leadership. You export six months of user engagement data from your analytics tool — 400 rows of event counts, retention rates, and feature usage by cohort. You paste it into this prompt and get back a structured briefing you can walk into the meeting with, instead of spending three hours in Excel.
Prompt
Act as a senior data analyst with 10+ years of experience in [YOUR INDUSTRY, e.g. e-commerce / SaaS / retail / finance]. Context: I have the following raw data from [DATA SOURCE, e.g. our monthly sales report / customer database export / marketing campaign results]: [PASTE YOUR DATA HERE — CSV rows, a table, or plain numbers are all fine] Analysis period: [DATE RANGE, e.g. January–May 2026] Business context: [1-2 sentences on what this data tracks and what decisions depend on it] Task: Analyze this data and produce a structured insight report. Go beyond surface-level observations — identify patterns, anomalies, correlations, and inflection points that a busy executive would not notice by scanning rows. Constraints: - Do not restate the raw data back to me - Every insight must include the specific number or percentage that supports it - Flag any data quality issues (gaps, outliers, inconsistencies) before drawing conclusions - Prioritize insights by business impact, not statistical novelty - Avoid jargon; write for a non-technical decision-maker Output Format: ## Executive Summary (3 bullet points max — the three things that matter most) ## Key Trends (numbered list — each trend in one sentence + supporting figure) ## Anomalies & Risks (anything that breaks the expected pattern and why it warrants attention) ## Opportunities (patterns that suggest an untapped action — be specific) ## Recommended Next Steps (3 concrete actions, each with a suggested owner and timeframe) ## Data Quality Notes (flag any gaps, inconsistencies, or caveats before conclusions are acted on)
Result
Executive Summary
• Feature adoption dropped 23% in March despite user count holding steady — activation, not acquisition, is the bottleneck.
• Power users (top 10% by session length) account for 61% of all feature interactions; casual users have barely touched core functionality after onboarding.
• Retention curve shows a sharp cliff at Day 14: 58% of users who reach Day 7 drop before Day 14, suggesting a specific friction point in the second week.
Key Trends
1. Daily active users grew 8.4% month-over-month from January to February, then plateaued — February's growth appears to be a one-time spike from the Product Hunt launch, not sustained organic growth.
2. Session duration increased 12% overall, but this is entirely explained by the power user segment; median session length for all other users is down 7% since the UI refresh in March.
3. Mobile users complete the onboarding flow at a 31% lower rate than desktop users (42% vs. 73%) — a consistent gap across all three months in the dataset.
Anomalies & Risks
• Week of April 14: a 34% single-week spike in support tickets correlates with the v2.3 release date. Not reflected in churn yet, but historically our churn lags support volume by 3-4 weeks — watch the May 5-12 cohort closely.
• Four enterprise accounts each show zero activity for 18+ days while their contracts are active. Possible churn signal; none flagged in CRM.
Opportunities
• Users who trigger the "Saved Report" feature within their first 10 days retain at 2.3x the rate of those who don't. It's used by only 19% of new users. A targeted onboarding nudge toward this feature could materially move 30-day retention.
• Mobile onboarding gap (42% vs 73%) represents a recoverable opportunity — if mobile completion reached desktop parity, projected 30-day MAU would increase by ~1,200 users at current acquisition rates.
Recommended Next Steps
1. Product: A/B test an onboarding prompt surfacing the Saved Report feature on Day 3. Owner: [PRODUCT LEAD]. Timeline: Sprint starting June 30.
2. Customer Success: Flag the four dormant enterprise accounts for outreach this week before renewal windows open. Owner: [CS MANAGER]. Timeline: By EOW.
3. Engineering: Audit the mobile onboarding flow for the specific drop-off step — Mixpanel funnel view will isolate it. Owner: [MOBILE LEAD]. Timeline: 1 week.
Data Quality Notes
• Rows 87-94 appear to be duplicate session events (same user ID, same timestamp ±2 seconds). Excluded from retention calculations.
• "Feature B" usage data is missing for March 1-7 — likely a tracking gap from the v2.2 deploy. Trends involving Feature B should be interpreted with caution for Q1 totals.
A maioria das equipes está afogada em exportações de dados que nunca analisam por completo. Este prompt funciona como um analista sênior sob demanda: você insere números brutos e ele devolve um relatório estruturado do jeito que executivos realmente leem — resumo executivo primeiro, anomalias sinalizadas, oportunidades identificadas e próximos passos atribuídos.
O prompt é construído em torno de uma estrutura de saída em cinco partes que força a IA a priorizar insights pelo impacto nos negócios, não pelo interesse estatístico. Também inclui uma seção obrigatória de qualidade de dados, para que você identifique lacunas e inconsistências antes que elas levem a decisões erradas.
O que torna este prompt diferente
A maioria dos prompts do tipo "analise meus dados" produz uma parede de estatísticas. Este instrui o modelo a pular a repetição dos dados, exigir números de apoio para cada afirmação e escrever para um tomador de decisão não técnico — o que significa que a saída é algo que você pode realmente encaminhar para a liderança sem editar.
Os campos [entre colchetes] permitem que você o personalize para qualquer setor ou fonte de dados. Insira um CSV de vendas, uma exportação de atribuição de marketing, uma tabela de engajamento de usuários ou até mesmo um resumo digitado manualmente de métricas semanais — a estrutura se adapta.
Quando usar
Este prompt conquista seu lugar em qualquer fluxo de trabalho recorrente de relatórios: revisões trimestrais de negócios, preparação para o board, reuniões semanais de equipe ou qualquer momento em que você recebe uma planilha e ouve "o que isso diz?". Funciona em diversos modelos — GPT-5, Claude Opus 4.8 e Gemini 2.5 Pro lidam bem com grandes entradas tabulares, embora Claude e Gemini tenham limites de contexto maiores para datasets muito grandes.
Dica profissional
Adicione uma linha ao seu contexto: "A decisão mais importante que depende desses dados é [DECISÃO]." Isso direciona o modelo para longe de observações genéricas e em direção ao insight específico que sua equipe realmente precisa.