User Prompts

# User Prompts for Akkio

## Prompt 1: Lead Scoring Model
"I've uploaded our CRM data with columns: lead_id, company_size, industry, lead_source (organic, paid, referral, event), website_visits, email_opens, email_clicks, demo_requested (yes/no), whitepaper_downloads, days_since_first_touch, annual_revenue_estimate, job_title, converted (yes/no). Build me a lead scoring model that: (1) Predicts which leads will convert, (2) Ranks all current open leads by conversion probability, (3) Shows me the top 5 most important features driving conversions, (4) Creates score buckets (Hot, Warm, Cool, Cold) with recommended actions for each, (5) Tells me the model accuracy and how confident we should be in these scores, (6) Identifies the ideal lead profile based on the patterns."

## Prompt 2: Revenue Forecasting
"Here's our monthly revenue data for the past 3 years: month, revenue, new_customers, churned_customers, marketing_spend, sales_headcount, product_launches (count), seasonality_index. Build a revenue forecast for the next 12 months: (1) Predict monthly revenue with confidence intervals, (2) Identify the key drivers of revenue growth, (3) Show the impact of increasing marketing spend by 20%, (4) Model three scenarios: conservative (reduce marketing 10%), baseline (current trend), aggressive (increase marketing 30% + hire 5 more salespeople), (5) Flag any months where the model predicts significant risk of decline, (6) Create a visual forecast chart with historical actual vs predicted."

## Prompt 3: Customer Churn Prediction
"My subscription data includes: customer_id, signup_date, plan_tier (basic/pro/enterprise), monthly_payment, usage_last_30_days, usage_trend (increasing/stable/decreasing), support_tickets_last_90_days, nps_score, contract_type (monthly/annual), days_until_renewal, feature_adoption_percentage, churned (yes/no). Build a churn prediction system: (1) Predict churn probability for each active customer, (2) Identify the top risk factors for churn, (3) Segment churned customers by reason (price, usage, support, competitor), (4) Calculate the expected revenue impact of predicted churn, (5) Recommend specific interventions for high-risk customers (discount, feature training, executive outreach), (6) Set up alerts for customers whose churn risk exceeds 70%."

## Prompt 4: Demand Forecasting for Inventory
"I have product sales data: product_id, product_category, date, units_sold, unit_price, promotion_active (yes/no), day_of_week, is_holiday, weather_temp, competitor_price, stock_level. Build a demand forecasting model for inventory planning: (1) Predict daily units sold per product for the next 30 days, (2) Factor in seasonal patterns, promotions, and holidays, (3) Calculate optimal reorder points and safety stock levels, (4) Identify products likely to stockout vs overstock, (5) Show the financial impact of forecast errors (cost of stockouts vs holding costs), (6) Create a weekly replenishment schedule recommendation."

## Prompt 5: Pricing Optimization
"My pricing experiment data includes: product_id, price_point_tested, units_sold, revenue, margin_percentage, customer_segment, channel (online/retail), competitor_price, time_period, elasticity_score. Help me optimize pricing: (1) Model the price-demand curve for each product, (2) Find the revenue-maximizing price point per product and segment, (3) Find the profit-maximizing price point (different from revenue-max), (4) Simulate the impact of a 10% price increase across the board, (5) Identify products where we're significantly underpriced vs overpriced relative to competitors, (6) Build a dynamic pricing recommendation engine that adjusts based on demand and competition."