{"id":"mkt2-first-vs-last-touch-attribution","audience":"marketing","category":"data","track":"marketing","technique":"verification","level":"advanced","title":"Count conversions by channel under first-touch and last-touch","free":false,"task":"You're comparing two attribution models for November's four converted customers. Write a prompt for a markdown table with columns Channel | First touch | Last touch listing Search, Social and Email with conversion counts, followed by a line starting \"Gains most under last touch:\" naming the channel and the change. First touch is the first channel in a customer's path and last touch the last. Ignore the visitor who didn't convert. Under 80 words, no other text.","materialLabel":"Customer paths (attached to your prompt automatically)","material":"NOVEMBER CONVERTED CUSTOMERS - touchpoint paths (earliest first)\n\nC1: Search > Email > Social > converted\nC2: Social > Search > converted\nC3: Email > Email > converted\nC4: Search > Social > Email > converted\n\nNOT CONVERTED\nV5: Social > Search (browsed only, never converted)\n\nChannels: Search, Social, Email.\nNote: the marketing lead says \"Search drives most conversions\". Only count converted customers. If a path lists the same channel twice, the channel is still counted once for that customer per model (first or last touch is a single channel).","rules":[{"type":"plainText"},{"type":"maxWords","max":80}],"mistakes":[{"title":"Counting the unconverted visitor","why":"The AI includes V5 in the totals.","fix":"Say: \"Only converted customers count.\"","checks":["r0","r2"]},{"title":"Mixing up the models","why":"The AI uses the middle touch or all touches.","fix":"Define: \"First touch = first channel in the path; last touch = last channel.\"","checks":["r0"]},{"title":"Following the marketing lead's view","why":"The AI says search drives most conversions.","fix":"Say: \"Use only the paths; no opinions.\"","checks":["r1"]},{"title":"Letting markdown through","why":"AI adds bold text and headings by default.","fix":"Add: \"Plain text only, no bold, no headings.\"","checks":["plainText"]},{"title":"Trusting the AI to count words","why":"AI is bad at counting and often runs past a word limit.","fix":"Cap each part (for example \"each line under 15 words\") so the total lands well under the limit.","checks":["maxWords"]}],"promptHints":[{"label":"Names the table columns","patterns":["channel.*first touch.*last touch"]},{"label":"Defines the two models","patterns":["first channel","last channel","first touch is","last touch"]},{"label":"Excludes the non-convert","patterns":["didn'?t convert","not convert","only converted","ignore"]},{"label":"Gives the closing line","patterns":["gains most under last touch:"]},{"label":"Caps length","patterns":["80 words","no other text"]}],"hint":"Attribution is counting with a rule. Define first and last touch, exclude non-converters and give the closing line format.","mission":"Count November conversions by channel, first touch versus last touch.","musts":["Ignore non-converter","Search, Social, Email","Name biggest last-touch gainer"],"rubricLabels":["The table has the columns Channel | First touch | Last touch with Search 2 and 1, Social 1 and 1, Email 1 and 2","The Gains most under last touch line names Email, up by 1","V5 (who didn't convert) is not counted, and there is no other commentary"]}