Analytics Prompts: 25 AI Prompts to Turn Data Into Decisions

"Help me define the right success metrics for my [BUSINESS TYPE]. I need to identify: 3-5 primary KPIs that directly indicate business health, 5-10 secondary metrics that support understanding, and vanity metrics I should ignore. For each metric, explain what it actually measures, when it's useful, and common misinterpretations."

"Audit my current analytics setup for [WEBSITE/APP]. I have [LIST CURRENT TRACKING]. Identify: what's working, what's broken or missing, data quality issues, and tracking gaps that could lead to wrong conclusions. Prioritize fixes by impact."

"For my [BUSINESS TYPE], map the relationships between these metrics: [LIST METRICS]. Show me: which metrics are leading vs lagging indicators, how they influence each other, which correlations matter vs which are spurious, and where I should focus for maximum impact."

"I have conflicting data between [PLATFORM A] and [PLATFORM B]. Help me: identify why they differ (tracking methodology, attribution windows, data processing), determine which is more reliable for what purposes, and create a reconciliation process to check data accuracy going forward."

"For my [INDUSTRY/BUSINESS TYPE], what are realistic benchmarks for: [LIST KEY METRICS]? Consider: industry averages, top quartile performance, what's achievable vs aspirational, and how benchmarks should change as we grow. Include context on why benchmarks vary."

"Design a conversion funnel for [BUSINESS TYPE] that tracks the user journey from first visit to purchase. Include 4-6 key events, identify where drop-offs typically happen, and suggest what data to capture at each step. Consider both micro-conversions and the final macro-conversion."

"Create an event tracking plan for [WEBSITE/APP]. Include: page view events, engagement events (scroll, video, downloads), conversion events, and custom events specific to my business. For each event, specify the event name, parameters to collect, and when it should fire."

"I have users moving between [DOMAIN 1] and [DOMAIN 2]. Help me set up cross-domain tracking to maintain session continuity, explain what configuration is needed in GA4 and GTM, and identify common mistakes that break the user journey tracking."

"Set up enhanced ecommerce tracking for my [PLATFORM]. Include: product impressions, product clicks, add to cart, remove from cart, begin checkout, add payment info, purchase, and refund events. Specify what data to pass with each event and common implementation pitfalls."

"Create a UTM tagging strategy for [BUSINESS] that covers: paid social, organic social, email campaigns, display ads, and referral traffic. Include naming conventions, when to use each parameter, a standard format for campaign names, and how to handle internal links vs external campaigns."

"Given this funnel data: [INSERT FUNNEL DATA], identify where the biggest drop-offs occur, calculate the conversion rate between each step, suggest why users might be dropping off at each point, and prioritize which step to fix first based on impact vs effort."

"Given this traffic source data: [INSERT DATA], analyze which sources bring the highest quality traffic. Consider: bounce rate, session duration, pages per session, conversion rate, and revenue per visitor. Identify which channels are overperforming vs underperforming and where to reallocate budget."

"Given this user behavior data: [INSERT DATA], identify natural audience segments based on behavior patterns, demographics, or acquisition source. For each segment, describe: their behavior, value to the business, and what messaging/UX would resonate. Suggest which segments to prioritize."

"Given this cohort retention data: [INSERT DATA], calculate retention rates for each cohort, identify patterns in when users drop off, compare retention across acquisition sources or user types, and suggest interventions to improve retention at critical drop-off points."

"Given this customer data: [INSERT DATA], calculate customer lifetime value by acquisition source, time to first purchase, repeat purchase rate, and average order value trends over time. Identify which acquisition channels bring the highest LTV customers vs. just the most customers."

"Design an executive dashboard for [STAKEHOLDER ROLE] that shows: 3-4 KPIs that matter most to their decisions, trend lines showing performance over time, alerts for metrics that need attention, and context that explains why numbers changed. Avoid vanity metrics and data that requires interpretation."

"Given this data: [INSERT DATA], write a performance report for [TIME PERIOD] that includes: what happened (the numbers), why it happened (analysis), what it means (implications), and what to do next (recommendations). Use plain language. Avoid data dumps. Lead with the most important insight."

"[METRIC] changed by [PERCENTAGE] in [TIME PERIOD]. Given this supporting data: [INSERT DATA], analyze what caused the change, rule out false explanations, identify the most likely root cause, and suggest how to validate your hypothesis. Consider seasonality, external factors, and data issues."

"I need to compare [PERIOD A] vs [PERIOD B] or [VARIANT A] vs [VARIANT B]. Create a comparison framework that accounts for: sample size differences, external factors, statistical significance, and practical significance. Include what metrics to compare, what conclusions are valid, and what comparisons are misleading."

"Design an alerts system for [BUSINESS TYPE] that monitors: critical metrics that should never drop, unusual patterns that indicate problems, opportunities that need fast action, and data quality issues. Include thresholds, who gets alerted, and what action to take for each alert type."

"I want to test [HYPOTHESIS] on [PAGE/ELEMENT]. Design an A/B test including: primary metric to optimize, secondary metrics to monitor, minimum sample size needed, test duration, what variants to test, and how to avoid common testing mistakes like peeking or multiple comparisons."

"Given these A/B test results: [INSERT RESULTS], determine if the result is statistically significant, calculate the confidence interval, assess practical significance (is the lift worth implementing?), identify segments where the variant performed differently, and recommend whether to implement, iterate, or abandon."

"Given this list of potential optimizations: [INSERT LIST], prioritize them using an ICE framework (Impact, Confidence, Ease). For each item, estimate the potential lift, how confident we are in that estimate, and the effort required. Return a ranked list with the top 3 priorities to focus on first."

"Given this customer journey data: [INSERT DATA], analyze which attribution model (first-touch, last-touch, linear, time-decay, position-based) best represents how our customers actually buy. Consider the typical sales cycle, number of touchpoints, and channel roles. Recommend a model and explain its limitations."

"Given this time-series data: [INSERT DATA], identify: unusual spikes or drops that need investigation, weekly/monthly seasonal patterns, trends that are accelerating or decelerating, and correlations between metrics. Flag anything that looks like a data quality issue vs. a real business change."