How do I do a root-cause drilldown to explain a change in a metric with iDBQuery?

A root-cause drilldown decomposes a change in a headline metric into the pieces that caused it, so a single number becomes an explanation. iDBQuery lets you drill through dimensions in plain language rather than writing layers of grouped SQL.

• You'd ask iDBQuery: 'Total revenue moved by £180k this month. Break that change down by product line, then region, then customer segment, and show which combinations contributed most.'
• It computes each dimension's contribution to the change, returns a contribution chart (a decomposition or waterfall), and lets you keep drilling.
• Every contribution is cited to its source rows, so 'the West region added £60k' traces to the actual records.

The strength is the interactive follow-up in one thread: 'Now drill the West region gain down by product,' then 'and by new vs returning customers.' iDBQuery keeps the whole context, so drilling never loses the thread. This handles both good and bad movements, and separates volume changes from mix shifts. Because the dimensions may span several sources, the one live model joins them so the decomposition is complete with no pipeline. For a hands-off version, the Analyst agent will pursue the drilldown itself, following the largest contributor at each level and reporting the root cause with the rows that prove it. This turns 'why did the number move?' into a verifiable, layered answer.