Sign in Get Started Free

The Library

Browse Every Prompt

Filter by platform, role, function or prompt type. Save the ones you want, copy them straight into your favourite AI tool.

59 prompts·Function: Data & Analytics

Build a spreadsheet formula from plain English

Plain-English description in, paste-ready spreadsheet formula out — Excel or Google Sheets dialect. Includes the "when it breaks" caveats most AI prompts skip.

Convert an Excel formula to Google Sheets (or vice versa)

Convert spreadsheet formulas between Excel and Google Sheets, with a callout of any function that doesn't exist in the target and the workaround used.

Outcome measure analysis agent

Quarterly review: takes your full caseload's outcome scores, identifies what's working, what's not, where you're an outlier (pos or neg) vs published benchmarks.

Workflow: monthly metric review prep

Prepares the monthly leadership metric review with anomaly investigation + action prompts.

Workflow: weekly product analytics review

Reviews weekly product analytics + recommends product changes.

Build a 'how to read a chart' guide

Drafts a guide on reading charts critically — don't be fooled by misleading visualisations.

Build a 'metric ownership' framework

Assigns clear ownership for every important metric — who's accountable for it.

Build a 'one number you stare at' practice

Picks the single number a team stares at every day — to focus action.

Build a 'data review meeting' for a team

Designs a recurring data review meeting for an individual team.

Build a 'dashboard request triage'

Triages dashboard requests from stakeholders to prevent dashboard sprawl.

Build a 'data trust' building protocol

Plans how a data team builds trust with stakeholders over time.

Build a 'is this number weird' check

Provides a quick check for whether a metric is anomalous before acting on it.

Build an experimentation playbook

Drafts an experimentation playbook covering ideation, design, run, learn.

Build an 'AB test results to action' protocol

Converts an A/B test result into a deployment + learning action plan.

Build a 'data democratisation' program

Plans a program to spread data literacy + access across the company.

Build a 'metric definitions diff' protocol

Catches when metric definitions silently drift over time and breaks comparisons.

Build an 'experiment ideas backlog' format

Drafts a format for collecting + prioritising experiment ideas across the team.

Build a cookieless attribution model

Designs an attribution model that works without third-party cookies.

Build a 'data > opinion' culture protocol

Designs a protocol for shifting from opinion-driven to data-informed decisions.

Build a privacy-friendly analytics setup

Sets up analytics that respects privacy + reduces consent fatigue.

Build a data dictionary

Drafts a data dictionary for a database or warehouse — the human-readable schema.

Build an analytics review meeting agenda

Drafts a recurring analytics review meeting agenda for the leadership team.

Build a 'data product' brief

Briefs a data product (dashboard, model, dataset) like a real product with users + outcomes.

Build an EDA (exploratory data analysis) plan

Plans an exploratory data analysis on a new dataset before forming hypotheses.

Build a 'metrics that matter' framework

Picks the 5 metrics that actually matter for a business stage and ignores the rest.

Run cohort analysis

Designs a cohort analysis with segmentation and visualisation choices.

Build an LTV prediction model

Designs a customer LTV prediction model with features and validation.

Run anomaly investigation

Investigates a metric anomaly with a structured root-cause approach.

Build a metric anomaly alerting plan

Sets up alerting on metrics with thresholds, channels, and noise discipline.

Run causal inference (diff-in-diff)

Plans a difference-in-differences analysis to test the causal effect of an intervention.

Run a survival analysis

Plans a survival analysis for time-to-event data like churn or upgrade time.

Build a self-serve analytics culture

Plans the shift from analyst-bottlenecked reporting to self-serve analytics.

Build a metric review meeting

Designs a recurring metric review meeting that drives action, not status.

Design an A/B test

Designs an A/B test with hypothesis, variants, primary metric, and decision rules.

Run a multivariate test

Designs a multivariate test where multiple elements vary at once.

Design a multi-armed bandit

Designs a multi-armed bandit experiment for ongoing optimisation rather than one-shot A/B.

Calculate statistical significance

Calculates significance for an A/B test result with confidence interval and interpretation.

Build a customer health score model

Designs a customer health score model with weighted signals and intervention triggers.

Run funnel analysis

Designs a funnel analysis with stages, drop-off identification, and recommendations.

Run engagement analysis

Plans an engagement analysis to understand depth + frequency of product use.

Build a churn prediction model

Designs a churn prediction model with features, validation, and intervention strategy.

Build a metric tree

Builds a metric tree decomposing a north star metric into its drivers.

Build a data quality framework

Designs a data quality framework with checks, owners, and incident response.

Build a data lineage doc

Documents data lineage from source to consumed metric for trust + debugging.

Build a data observability plan

Plans data observability with monitoring of pipelines, freshness, and quality.

Build a 'correlation vs causation' check

Provides a checklist for distinguishing correlation from causation before acting on data.

Build a 'metric definition' doc

Drafts a precise metric definition that prevents 'we're talking about different things' confusion.

Calculate sample size

Calculates the sample size needed for a survey or A/B test with confidence.

Run sentiment analysis on customer feedback

Codes customer feedback for sentiment with positive, negative, and mixed signal classification.

Run topic modelling on text data

Plans topic modelling (LDA or BERTopic) on a corpus of text data.

Run regression analysis

Plans a regression analysis to test the relationship between variables.

Run cluster analysis

Plans a cluster analysis to find natural groupings in customer data.

Build a forecast model

Designs a forecast model for revenue, headcount, or other future metrics.

Write a marketing performance dashboard spec

Specifies the metrics, layout, and cadence for a weekly marketing performance dashboard.

Write a quarterly social media report

Drafts a quarterly social media performance report with insights and next-quarter recommendations.

Build an attribution model framework

Helps a small team choose and implement an attribution model that fits their stack and budget.

Build a CAC by channel analysis

Calculates customer acquisition cost by channel and recommends where to invest more or less.

Build a churn forensics report

Diagnoses churn root causes with cohort, segment, and reason analysis.

Build an executive dashboard spec

Specifies the 8 metrics and layout for a CEO/leadership dashboard reviewed weekly.