Seasonal Sales → Campaign Planning

Use Shopify historical order data with Mavera to generate data-driven seasonal campaign calendars

Scenario

Your store has years of order history with seasonal patterns — holiday spikes, summer slowdowns, back-to-school surges. You pull historical orders, aggregate revenue and order count by month, then feed that data to Mave Agent for a campaign calendar with timing, messaging themes, and budget allocation.

Architecture

Code

Python

import os, requests
from collections import defaultdict

STORE = os.environ["SHOPIFY_STORE"]
TOKEN = os.environ["SHOPIFY_ACCESS_TOKEN"]
MV = os.environ["MAVERA_API_KEY"]
SH = f"https://{STORE}.myshopify.com/admin/api/2024-10/graphql.json"
SH_H = {"X-Shopify-Access-Token": TOKEN, "Content-Type": "application/json"}
MB = "https://app.mavera.io/api/v1"
MH = {"Authorization": f"Bearer {MV}", "Content-Type": "application/json"}

QUERY = """query ($cursor: String) {
  orders(first: 250, after: $cursor, sortKey: CREATED_AT, query: "created_at:>2023-01-01") {
    edges { node { createdAt totalPriceSet { shopMoney { amount } } lineItems(first: 3) { edges { node { title quantity } } } } }
    pageInfo { hasNextPage endCursor }
  }
}"""

orders, cursor = [], None
while True:
    resp = requests.post(SH, json={"query": QUERY, "variables": {"cursor": cursor} if cursor else {}}, headers=SH_H)
    resp.raise_for_status()
    data = resp.json()["data"]["orders"]
    orders.extend(e["node"] for e in data["edges"])
    if not data["pageInfo"]["hasNextPage"]: break
    cursor = data["pageInfo"]["endCursor"]
monthly = defaultdict(lambda: {"revenue": 0, "orders": 0, "items": defaultdict(int)})
for o in orders:
    k = o["createdAt"][:7]
    monthly[k]["revenue"] += float(o["totalPriceSet"]["shopMoney"]["amount"])
    monthly[k]["orders"] += 1
    for li in o["lineItems"]["edges"]:
        monthly[k]["items"][li["node"]["title"]] += li["node"]["quantity"]
lines = ["Month | Orders | Revenue | AOV | Top Product", "------|--------|---------|-----|------------"]
for m, d in sorted(monthly.items()):
    top = max(d["items"], key=d["items"].get) if d["items"] else "N/A"
    lines.append(f"{m} | {d['orders']} | ${d['revenue']:.0f} | ${d['revenue']/d['orders']:.0f} | {top}")
summary = "\n".join(lines)
print(summary)
plan = requests.post(f"{MB}/mave/chat", json={"input": [
    {"role": "system", "content": "Senior marketing strategist. Given monthly sales data, create a 12-month campaign calendar. Per campaign: timing, theme, audience, messaging, budget share (%)."},
    {"role": "user", "content": f"Sales data:\n\n{summary}\n\nCreate campaign calendar for next 12 months."},
]}, headers=MH)
plan.raise_for_status()
print(f"\n--- Campaign Plan ---\n{plan.json()['choices'][0]['message']['content']}")

Example Output

Month | Orders | Revenue | AOV | Top Product
------|--------|---------|-----|------------
2024-01 | 285 | $34200 | $120 | Merino Base Layer
2024-11 | 487 | $72100 | $148 | Down Expedition Parka
2024-12 | 623 | $99680 | $160 | Gift Card Bundle

--- Campaign Plan ---
## Q4: Peak Season (Nov–Dec) — 40% budget
- Nov 1–15: Early bird holiday sale — VIP first access
- BFCM: Best-sellers with bundle offers
- Dec 1–20: Gift guide segmented by price range

Error Handling

Query cost for large order histories

Fetching 250 orders with line items costs 20–30 query points per request. Monitor extensions.cost.throttleStatus.currentlyAvailable. For 50k+ orders, filter by quarterly date ranges or use Shopify Bulk Operations.

Mave Agent response truncated

If the summary exceeds the context window, aggregate quarterly instead of monthly and limit to top 3 products per period. Split into two calls — pattern identification, then campaign planning.