Post Purchase Survey Trigger

Read-only: identifies orders 7–14 days post-fulfillment that are eligible for a post-purchase survey campaign, excluding refunded or cancelled orders.

shopify-admin-post-purchase-survey-trigger


Purpose

Builds the recipient list for a post-purchase survey campaign by selecting orders that were fulfilled between survey_min_days and survey_max_days ago, are not refunded, not cancelled, and (optionally) belong to customers who consented to marketing. Output is a clean recipient list ready to load into your email or SMS automation. Read-only — no mutations.


Prerequisites

  • Authenticated Shopify CLI session: shopify store auth --store --scopes read_orders,read_customers,read_fulfillments
  • API scopes: read_orders, read_customers, read_fulfillments

  • Parameters


    ParameterTypeRequiredDefaultDescription
    storestringyesStore domain (e.g., mystore.myshopify.com)
    survey_min_daysintegerno7Earliest days after fulfillment to survey (give time for delivery + initial use)
    survey_max_daysintegerno14Latest days after fulfillment to survey (recall fades after ~2 weeks)
    marketing_consent_onlyboolnotrueRestrict to customers with marketingState: SUBSCRIBED
    exclude_repeat_recipients_daysintegerno90Skip customers who were already on a survey list within this window (caller-tracked)
    formatstringnohumanOutput format: human or json

    Safety


    > ℹ️ Read-only skill — no mutations are executed. The skill produces a recipient list; the caller is responsible for actually sending the survey via their own email / SMS platform. Honor marketing_consent_only: true for promotional surveys to stay compliant with consent rules.


    Workflow Steps


  • Compute window: latest_fulfilled_at = NOW - survey_min_days, earliest_fulfilled_at = NOW - survey_max_days

  • OPERATION: orders — query
  • Inputs: query: "fulfillment_status:fulfilled financial_status:paid -status:cancelled updated_at:>=", first: 250, select displayFulfillmentStatus, displayFinancialStatus, cancelledAt, fulfillments { createdAt, displayStatus, deliveredAt }, refunds { id, createdAt }, customer { id, defaultEmailAddress { emailAddress, marketingState }, displayName, locale }, pagination cursor

    Expected output: All candidate orders; paginate until hasNextPage: false


  • Filter orders to the survey window using the earliest fulfillment createdAt:
  • Skip if fulfilled_at is outside [earliest_fulfilled_at, latest_fulfilled_at]
  • Skip if cancelledAt != null
  • Skip if any refund was issued (full or partial)
  • Skip if customer is null (guest order without email)
  • If marketing_consent_only: true, skip if marketingState != SUBSCRIBED

  • De-duplicate on customer email — multiple orders from the same customer collapse to a single survey invite

  • Output recipient list

  • GraphQL Operations


    # orders:query — validated against api_version 2025-01
    query SurveyEligibleOrders($query: String!, $after: String) {
      orders(first: 250, after: $after, query: $query) {
        edges {
          node {
            id
            name
            createdAt
            cancelledAt
            displayFinancialStatus
            displayFulfillmentStatus
            totalPriceSet {
              shopMoney {
                amount
                currencyCode
              }
            }
            fulfillments(first: 5) {
              id
              createdAt
              displayStatus
              deliveredAt
            }
            refunds {
              id
              createdAt
            }
            customer {
              id
              displayName
              locale
              defaultEmailAddress {
                emailAddress
                marketingState
              }
            }
          }
        }
        pageInfo {
          hasNextPage
          endCursor
        }
      }
    }
    

    Session Tracking


    Claude MUST emit the following output at each stage. This is mandatory.


    On start, emit:

    ╔══════════════════════════════════════════════╗
    ║  SKILL: Post-Purchase Survey Trigger         ║
    ║  Store: <store domain>                       ║
    ║  Started: <YYYY-MM-DD HH:MM UTC>             ║
    ╚══════════════════════════════════════════════╝
    

    After each step, emit:

    [N/TOTAL] <QUERY|MUTATION>  <OperationName>
              → Params: <brief summary of key inputs>
              → Result: <count or outcome>
    

    On completion, emit:


    For format: human (default):

    ══════════════════════════════════════════════
    POST-PURCHASE SURVEY LIST
      Window:                 <survey_min_days>–<survey_max_days> days post-fulfillment
      Orders considered:      <n>
      Refunded (excluded):    <n>
      Cancelled (excluded):   <n>
      No consent (excluded):  <n>
      Guest orders (excluded):<n>
      ─────────────────────────────
      Eligible recipients:    <n>
      Output: post_purchase_survey_<date>.csv
    ══════════════════════════════════════════════
    

    For format: json, emit:

    {
      "skill": "post-purchase-survey-trigger",
      "store": "<domain>",
      "window_min_days": 7,
      "window_max_days": 14,
      "orders_considered": 0,
      "excluded": { "refunded": 0, "cancelled": 0, "no_consent": 0, "guest": 0 },
      "eligible_recipients": 0,
      "output_file": "post_purchase_survey_<date>.csv"
    }
    

    Output Format

    CSV file post_purchase_survey_.csv with columns:

    customer_id, email, name, locale, order_name, order_total, currency, fulfilled_at, delivered_at, days_since_fulfilled


    Error Handling

    ErrorCauseRecovery
    THROTTLEDAPI rate limit exceededWait 2 seconds, retry up to 3 times
    Order with multiple fulfillmentsSplit shipmentUse earliest createdAt for window calc; surface in CSV as multi_fulfillment: true
    deliveredAt nullCarrier didn't report deliveryUse fulfillments.createdAt as proxy (may slightly under- or over-shoot)
    Customer with no emailPhone-only profileSkip — survey campaigns assume email channel

    Best Practices

  • Send the survey 7–14 days after fulfillment: 7 days is enough for delivery + first use, 14 days is the sweet spot before recall decays.
  • For high-cost or considered-purchase categories (furniture, electronics), extend survey_max_days to 30 — customers form opinions later.
  • Pair with customer-cohort-analysis to compare survey-respondent retention vs. non-respondents — engaged respondents are typically your stickiest cohort.
  • Never include customers with refunded or cancelled orders — surveying them invites complaints, not feedback you can act on.
  • Run this skill on a daily schedule and feed the CSV directly into your email automation; manual cadence ruins the time-window targeting.