Implementation Strategy and Variable/Tag Design
Variable design is the discipline of making data usable years later. The candidate must know Adobe Analytics variable types and their persistence: eVars (conversion variables with configurable allocation and expiration), props (traffic variables, typically hit-scoped), events (counters, currency, numeric), list vars and merchandising eVars for multi-value and product-level attribution, and processing-time dimensions such as classifications and marketing channels. Understand allocation (last, first, linear, most recent) and expiration (hit, visit, purchase, event, never) as the primary levers that determine what a report actually answers.
Tag design covers the tagging library structure: rules, conditions, data elements, extensions, and the data layer as the single source of truth. Know why direct DOM scraping is fragile, why data element reuse enforces naming consistency, and how rule ordering and load order affect whether values are present at send time.
Traps: setting conversion-variable expiration to 'never' on high-cardinality values; using props for values that must persist across a visit; relying on default allocation when the business question implies first-touch; duplicating the same concept in two variables; and using a merchandising eVar without the required product syntax. Questions often present a plausible variable that technically works but produces misleading attribution.
Rule of thumb: design each variable around one question and one cardinality budget. If a value can exceed a few thousand distinct entries per month, it belongs in a classification or a report-suite-external store, not a raw eVar.
Worked example: an internal search term must attribute revenue to the last term used before purchase. Use an eVar with last allocation and purchase expiration, not a prop, because hit scope loses the term at conversion.