Read a survey without fooling yourself
Here are survey results: {{paste questions, response counts, and open-text answers}}. It was distributed via {{channel}} to {{who}}, with {{n}} responses out of {{invited}}. 1. St…
Data & Analysis prompts
Practical data & analysis prompts tested with useful AI apps.
Unit economics that survive scrutiny
Work out the unit economics for {{business}}. Inputs: {{price, COGS components, CAC by channel, churn, expansion, gross margin, payback expectations}}. 1. Define each metric expli…
Size a market from the bottom up
Size the market for {{product}} sold to {{customer type}} in {{geography}}. Build it bottom-up, not from an analyst's TAM number: 1. Define the unit of demand precisely (a company…
Cohort retention analysis in plain language
Here's my signup and activity data: {{describe columns or paste sample}}. 1. Build monthly signup cohorts and show retention by month-since-signup as a triangle table. State clear…
Write the query and the caveat for a stakeholder question
A stakeholder asked: "{{their question, verbatim}}". Available data: {{schema summary}}. 1. Restate what they're actually trying to decide — the question behind the question. If t…
Clean a messy spreadsheet reproducibly
Here's a messy export: {{paste 20-30 sample rows with headers}}. 1. Diagnose every problem you can see: inconsistent date formats, numbers stored as text, mixed units, trailing wh…
Interrogate a dataset before you trust it
Here's a dataset: {{paste sample rows plus column list, row count, and how it was collected}}. Before any analysis, audit it: 1. For each column: type, unit, expected range, % mis…
Design an experiment that can actually conclude
I want to test {{change}} on {{surface}}. Baseline: {{metric and current value}}, traffic {{volume per week}}. Design the test: 1. State the hypothesis as a mechanism, not a hope:…
Explain a metric move without inventing a story
{{Metric}} moved from {{before}} to {{after}} between {{dates}}. Context: {{what you know — launches, campaigns, seasonality, outages}}. Data available: {{list dimensions you can s…
Find the duplicate records ruining your counts
My count is wrong: {{what you expected versus what you got}}. Here's the data: {{schema and 15 sample rows, including a few you suspect}}. Work through duplicate types in this ord…
Forecast with the uncertainty attached
Forecast {{metric}} for the next {{horizon}}. History: {{paste time series}}. Known future events: {{launches, seasonality, pricing changes}}. 1. Describe the series honestly firs…
Pick the right chart for the point you are making
Here's my data: {{describe the columns, the row count, and what each row represents}}. The point I want to land: {{the one sentence a reader should take away}}. Audience: {{who}}. …
Build the dashboard someone will actually use
Design a dashboard for {{role}} who needs to decide {{decision}} on a {{cadence}} basis. Available data: {{list}}. Start from decisions, not charts: 1. List the decisions this per…
Explain a statistical result to a non-technical exec
Here's the result: {{paste the numbers — effect size, confidence interval, p-value, sample size, whatever you have}}. Method: {{what you did}}. Audience: {{role}}, who will use it …
Turn analysis into an executive memo
Here's my analysis: {{paste findings, charts described, numbers}}. Audience: {{who}}, who cares about {{their objective}} and has 4 minutes. Write a one-page memo: 1. BOTTOM LINE …
Estimate what a price change will really do to revenue
You are a pricing analyst. I am considering the price change described below. Using only my data, build the revenue case honestly. Output: 1. BASELINE — current customers, ARPU, …
Turn a churn export into the three cancel reasons you can act on
You are a retention analyst. Here is an export of customers who cancelled: {{paste rows — signup date, cancel date, plan, price, usage in the last 30 days, stated cancel reason, an…
Find the story in a metric that moved
You are a data analyst. A metric moved and I need to know why before I report it. I will give you the metric, the period, the change, and any breakdowns I have. Work in this orde…
Reconcile two reports that disagree about the same number
Two sources report different values for the same metric and I have to explain the gap today. Metric: {{name}}. Source A: {{tool, value, how it is defined if you know}}. Source B: {…
Decide which numbers belong on a dashboard nobody asked to see
You are an analyst cutting a bloated dashboard down to what changes decisions. Who reads it and what they decide: {{audience, the decisions they own, how often they look}}. Current…
Write the data dictionary nobody wants to write
Here's a table: {{paste the schema, plus 10-20 sample rows}}. Team context: {{what the table is used for, who queries it}}. For every column produce: the plain-English meaning, th…
Work out how much of last month's growth was real
You are a skeptical analyst. Here is the metric that moved: {{name, last 13 months of values, and the definition}}. What else changed: {{launches, pricing, campaigns, tracking chan…
Decide the sample size before you run the test
I want to test {{change}} on {{surface}}. Current baseline: {{metric and its current rate or mean, plus its variance if you have it}}. Traffic: {{visitors or events per day eligibl…
Sanity-check a spreadsheet model before you present it
You are a skeptical reviewer. Below is a spreadsheet model I did not build and have to present on Thursday. Find what would embarrass me. Return: 1. ASSUMPTIONS — table with Cell…
Segment your customers by what they are actually worth to you
You are an analyst who knows revenue and value are not the same number. Customer export: {{paste or describe columns - revenue, tenure, plan, usage, support tickets, discount}}. W…
Turn a support ticket export into the five fixes worth making
You are a support operations analyst. Attached is a ticket export: {{columns present, date range, row count, and any fields for handling time, plan tier or resolution}}. I want to …
Turn a raw event log into a funnel with the drop-off explained
You are a product analyst. I have raw event data and need a funnel I can defend. Event schema: {{table name, columns, event names, and what each one fires on}}. The journey I care …
Work out whether a dip is seasonal or something breaking
You are an analyst who refuses to call something seasonal without checking. My metric dropped and half the team says it is the time of year. Below is the history and what changed. …
Work out which customers are quietly about to leave
You are a data analyst who has seen too many churn models that only fire after the customer has already gone. The export: {{paste usage or account data - columns and 20+ rows, or …
Decide whether an A/B test result is safe to ship
You are a sceptical experimentation reviewer. I want to ship a test result and need to know if it holds. The test: {{what changed, primary metric, hypothesis}}. Numbers: {{per vari…
Work out what your data can honestly answer before you promise an analysis
You are a senior analyst scoping a request before committing to it. Half of these questions cannot be answered with what we actually collect, and finding that out in week two is th…
Work out whether a broken dashboard number is data or reality
You are a data lead. A number on our dashboard looks wrong. Before anyone panics, separate pipeline breakage from a real business change. Output four sections. 1. FIRST FIVE CHECK…
Decide what to instrument before you ship the feature
You are a product analyst reviewing a tracking plan before code freeze. Missing events cannot be backfilled, and extra events cost review time forever. The feature and how a user …
Find the cost lines worth cutting in a year of spend
You are a finance analyst. Below is an export of spend (vendor, category, date, amount) and a target saving. Find the cuts, not a generic cost-cutting essay. 1. SHAPE OF SPEND - t…
Work out whether a new feature changed behaviour or just moved it around
You are an analyst checking whether a shipped feature created new activity or cannibalised existing activity. Below is usage data before and after launch, plus what the feature doe…
Set the alert thresholds that catch a real problem, not noise
You are an analyst setting up alerting on business and product metrics. Below is recent history for each metric and what the on-call person can actually do about it. Design alerts …
Decide what to do with the outliers before they decide for you
You are a careful analyst. Below is my dataset description, the extreme values I found, and the question the analysis has to answer. Decide how each outlier should be handled. 1. …