How to use this library
Replace placeholders like [PRODUCT], [CATEGORY], [MARKET] with whatever you're curious about. Wayflyer Intelligence will ask follow-ups when it needs to.
1. Getting Started: First Look
The fastest way to see what Wayflyer Intelligence can do. Great for your first session.
"How is my business performing this week?"
"What were the top 3 highlights and 3 concerns from last month?"
"Give me a snapshot of this month compared to last month"
"Prepare a short summary for my next board meeting"
2. Sales & Revenue
Diagnose revenue trends, market splits, AOV shifts.
"Break down my revenue by sales channel for the last 30 days"
"How are my sales trending this month vs the same period last year?"
"Why has my average order value changed recently?"
"Compare revenue between my UK, US and ROW markets for this quarter"
"Which day of the week generates the most revenue?"
"Where did the order shortfall in [MONTH] come from?"
3. Products
Top, bottom and edge cases across your catalogue. Product-level only.
"What are my top 10 products by revenue this month?"
"Which products grew the most in units sold YoY?"
"Are there products I should consider dropping from my range?"
"Show me products with strong revenue but high discounting"
4. Marketing & Ads
Channel and campaign efficiency, with last-click attribution caveats.
"What's my ROAS by channel for the last 30 days?"
"Which marketing channel is most efficient at acquiring new customers?"
"How much am I spending on ads as a percentage of revenue?"
"Compare my Google Ads and Meta Ads performance this month"
"Which campaigns drove the most revenue last month?"
"Where should I be optimising my ad spend right now?"
5. Customers
New, returning, and where they're coming from.
"How many new customers did I acquire last month vs the same month last year?"
"What share of my revenue this quarter came from returning customers?"
"What's the average time between a customer's first and second order?"
"Which countries are my new customers coming from this month?"
"What's my repeat purchase rate this quarter?"
6. Returns & Refunds
Refund patterns and problem products.
"Which products had the highest refund rates in the last quarter?"
"How has my overall refund rate trended over the past 6 months?"
"Are refunds concentrated in any particular sales channel or country?"
7. Promotions & Discounts
Where discounting is helping and where it's just leaking margin.
"How much of my revenue this quarter went through promo codes?"
"Which promo codes drove the most orders in the last month?"
"Has my average discount rate increased compared to last year?"
8. Strategic Deep-Dives
CFO-style questions. Diagnostic and advisory.
"Where is my biggest area of underperformance right now?"
"What patterns do you see in my data that I should act on?"
"What would you focus on to grow revenue by 10% this year?"
"What's the biggest risk facing my business right now?"
"Tell me the story of my Q1. What went well and what didn't?"
9. Seasonal & Year-on-Year
Holiday and cyclical patterns.
"How did Black Friday 2025 compare to 2024 across orders, AOV and channel mix?"
"What can I learn from last summer to plan for this one?"
"Show me month-by-month revenue for 2025 vs 2026 so far"
"Which products had the strongest seasonal swings last year?"
Guidelines on Building Custom Prompts
Be specific about time periods. Use concrete date ranges rather than vague terms like "recently." The system resolves relative dates, but explicit ranges reduce ambiguity and get faster results.
Avoid asking for forecasts or numerical projections. The system will refuse precise predictions. Instead ask for trend analysis or patterns: "is my retention trending worse?" rather than "what will my Q4 revenue be?"
Use plain business language. The system translates business questions into data queries internally. You don't need to know metric names or tool syntax.
Ask follow-up questions to drill down. The system is designed for multi-turn conversation and will suggest next steps. Follow those threads rather than asking one giant compound question.
Expect attribution caveats on paid social. If you ask about Meta or TikTok performance, the system will always lead with attribution context before numbers. That's by design, not evasion.
Keep the data request and output format in one sentence. Rather than "pull customer revenue and put it in a table," say "show me customer revenue for the last 90 days in a table." Connecting what you want fetched with how you want it displayed gets cleaner, more reliable results.