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 <domain> --scopes read_orders,read_customers,read_fulfillments
• API scopes: read_orders, read_customers, read_fulfillments

Parameters

Parameter	Type	Required	Default	Description
store	string	yes	—	Store domain (e.g., mystore.myshopify.com)
survey_min_days	integer	no	7	Earliest days after fulfillment to survey (give time for delivery + initial use)
survey_max_days	integer	no	14	Latest days after fulfillment to survey (recall fades after ~2 weeks)
marketing_consent_only	bool	no	true	Restrict to customers with marketingState: SUBSCRIBED
exclude_repeat_recipients_days	integer	no	90	Skip customers who were already on a survey list within this window (caller-tracked)
format	string	no	human	Output 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

1. Compute window: latest_fulfilled_at = NOW - survey_min_days, earliest_fulfilled_at = NOW - survey_max_days

2. OPERATION: orders — query Inputs: query: "fulfillment_status:fulfilled financial_status:paid -status:cancelled updated_at:>=<earliest_fulfilled_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

3. 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

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

5. 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_<YYYY-MM-DD>.csv with columns: customer_id, email, name, locale, order_name, order_total, currency, fulfilled_at, delivered_at, days_since_fulfilled

Error Handling

Error	Cause	Recovery
THROTTLED	API rate limit exceeded	Wait 2 seconds, retry up to 3 times
Order with multiple fulfillments	Split shipment	Use earliest createdAt for window calc; surface in CSV as multi_fulfillment: true
deliveredAt null	Carrier didn't report delivery	Use fulfillments.createdAt as proxy (may slightly under- or over-shoot)
Customer with no email	Phone-only profile	Skip — 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.