Funnel reports exist for every stage, yet the stage boundaries were drawn at different times, so the numbers do not chain together. In this case, the immediate task is to analyze losses across the store conversion funnel. The workflow should produce a conversion diagnosis and experiment backlog, with each step count traced back to the merchant's own records.
First replace every [bracketed] placeholder with real store scope and constraints, then add any actual records or data you have, and paste it into your preferred AI assistant or Spark agent. Spark will follow this page's workflow to produce structured output and verification steps.
Act as a Shopify operations specialist.
Help me analyze losses across the store conversion funnel.
My store context is [describe the business, products and target audience].
My scope is [specify the affected products, pages, customers or reporting period]. Use these inputs: [each funnel step with its entry event and session rule, the count per step over one window, the definition, tool and boundary date recorded for each step, the segment splits required, and the order export for the same window].
Check that consecutive steps chain, meaning a visitor counted in a later step is also counted in every earlier step and the counts never rise going down the funnel, express each fall as a share of the preceding step, and mark as non-comparable any pair of steps whose definitions or boundary dates differ.
Return a conversion diagnosis and experiment backlog with every step count reconciled against the supplied records for the specified task.
Show the relevant evidence or before-and-after examples and explain how I can verify the result.
If essential data is missing, ask for it and mark unsupported conclusions as unknown.
Do not invent metrics, product claims or customer facts.
For any publication, message, account or store change, use only connected tools within my explicitly authorized scope and report success only after verification; otherwise provide the draft or proposed steps.Split the work into input, judgment, and deliverable phases. Verify each step's output before moving to the next.
Write down each funnel step with the event and the session rule that puts a visitor into it, then pull the count for every step over one window from the same source where possible, recording the definition, the tool and the date each step boundary was set, the segment splits required, and the order export for the same window.
Check that consecutive steps chain, meaning a visitor counted in a later step is also counted in every earlier step and the counts never rise going down the funnel, express each fall as a share of the preceding step, and mark as non-comparable any pair of steps whose definitions or boundary dates differ.
Deliver a conversion diagnosis and an experiment backlog naming the step with the largest fall, its share, the conditions recorded at that step and the proposed test, then reconcile the final step count with the merchant's order export for the window and list every step whose count could not be traced back to a record.
Cover decision rules, frequent misunderstandings, and required inputs so the team aligns on prerequisites before execution.
Two steps are comparable when both count the same visitors under one session rule over one window, and neither count comes from a source with a different attribution rule. Where a step boundary was set on a different date, that pair is marked non-comparable and given raw counts, because a definition change looks like a change in behaviour.
A visitor reaching a later step must also appear in every earlier step, so the counts can only stay level or fall as the funnel narrows. A rise between consecutive steps means the two count different populations or windows; that pair is flagged and redefined before any loss is calculated, otherwise the loss figure describes nothing real.
Only where one tool records every event the funnel needs, and that is checked step by step rather than assumed. Where a step lives in a second system, the join key, the window and the timezone are recorded with the count; steps that cannot be joined are reported as unmeasured so the funnel is not completed with an estimate.
Start with scenarios under the same topic, then expand into cross-topic matches suggested by similarity.
Ciwi Spark builds Shopify-native AI tools and agents around the workflows merchants actually run: research, ads, translation, and customer support. Stop rewriting the same prompts by hand.