Breaking down complex queries into multiple sentences is essential for improving interpretation, readability, and accuracy. Writing queries in a structured, step-by-step manner ensures better comprehension when revisiting saved insights and allows for more complex scenarios to be handled effectively. Single-sentence queries can sometimes be misinterpreted or incorrectly structured, leading to inaccurate results or no output at all. By breaking a query into multiple steps, the model can process the request more efficiently and interpret the logic more clearly.
For example, instead of writing a single long query, structuring it as "Show the average LTV by country and btag. Show LTV to 0 decimal places." ensures clarity and improves query execution.
Other examples
Show btags where total signups this month is lower than last month MTD. Include total signups this month and total signups last month MTD.
List countries with a total casino wager value greater than 1000 last month. Include total casino NGR, total wagers and total bonuses for the same period.
Show the top performing btags by conversion rate where country is GB. Show conversion rate to 2 decimal places as a percentage.
Show average RTP for games with a game category [Slots]. Only include data for last year. Only include data for players from the UK market.
Text enclosed within [ ] is not interpreted as a keyword, ensuring it does not affect query processing.