Missing Revenue in GA4

If you’re seeing GA4 report $0 or missing revenue for purchase events, here are the most common causes and how to fix them. The purchase event is one of GA4’s strictest events — it has specific required fields, strict type expectations, and a deduplication layer that silently drops hits. A single malformed field is all it takes for revenue to vanish.

1. `value` parameter missing or not mapped

Verify. GA4 → DebugView → inspect a recent purchase event. Look for the value parameter. If it’s absent, GA4 has no total-revenue number to attribute. The event counts (as a conversion), but revenue stays at 0.

Fix. Open the GA4 Event tag in GTM for purchases. Under Event Parameters, add value mapped to your Data Layer Variable (typically {{DLV - ecommerce.value}} or similar). Ensure the push sets ecommerce.value at the correct level — GA4 reads it from the top-level value, not from inside items.

2. `value` is a string instead of a number

Verify. BigQuery query: SELECT event_name, ecommerce.purchase_revenue FROM ... WHERE event_name = 'purchase' LIMIT 100. If purchase_revenue is NULL but raw event_params show value = "89.99" (string), GA4 rejected the value at ingest.

Fix. Cast to number in the dataLayer push. value: parseFloat(order.total) not value: order.total.toString(). In GTM, Data Layer Variables return whatever type the push sent — if the source is a string, either fix the source or use a JavaScript Variable to parse: return parseFloat({{DLV - ecommerce.value}});. Verify type with typeof in DevTools Console.

3. Missing or wrong `currency` code

Verify. currency must be a valid ISO 4217 three-letter code (USD, EUR, GBP, JPY, SEK, etc). "$", "dollar", "us", blank, or lowercase "usd" (valid but inconsistent) can cause rejection or misattribution. GA4 uses the property’s default currency when currency is missing, which may convert your values unexpectedly.

Fix. Always send currency in uppercase ISO 4217. For multi-currency stores, send the transaction’s currency (not the property’s base currency) and let GA4 do the conversion. Pin the value in your dataLayer spec — e.g. currency: 'EUR', not currency: 'eur' or currency: '€'.

4. `items` array missing prices

Verify. Even with value correct at the event level, missing items[n].price breaks item-level revenue reports. Look at GA4 → Monetisation → Item-level reports — if product-level revenue is 0 but top-line revenue is fine, this is the symptom.

Fix. Every item in items needs item_id, item_name, price (as a number), and quantity. price is per-unit, not the line total. If you’re sending price: "0" because the backend sends prices in cents as strings, fix it: price: product.price_cents / 100.

5. Tax and shipping inclusion mismatch

Verify. Is your value supposed to include tax and shipping or not? GA4’s convention is that value represents what GA4 should treat as revenue — which for most e-commerce is ex-tax, ex-shipping (the product revenue). If your backend sends value = order.grand_total (inc. tax + shipping) but your reporting assumes product revenue, GA4 “revenue” will be inflated vs. finance’s revenue.

Fix. Pick a convention with finance and stick to it. The cleanest pattern: value = order.subtotal (products only), plus separate tax and shipping parameters. GA4 receives all three and analysts choose which to report on. Document the decision.

6. Duplicate purchase deduplicated — wrong one won

Verify. If duplicate purchases (see Duplicate Events in GA4) land with the same transaction_id, GA4 dedups — keeping one, dropping others. If the surviving copy has value: 0 while the dropped one had the real amount, revenue disappears.

Fix. Pin down the duplicate source and kill one side. Also check which side sends complete data — a common pattern is client-side sends a rough estimate (cart total) and server-side sends the real settled amount. If GA4 keeps the client-side copy, you lose the good data. Ensure both copies have identical value, or remove the duplicate entirely.

7. Refund events subtracting from revenue

Verify. GA4 → Monetisation → compare Purchase revenue to Gross revenue. If Purchase revenue is noticeably lower and you have a flow that sends refund events, refunds are being subtracted. Check DebugView for refund events — especially ones that didn’t correspond to real refunds.

Fix. The refund event is designed to subtract. That’s correct. But ensure:

You only fire refund on actual refunds, not cancellations (unless the payment was captured and needs to be undone).
refund events include a transaction_id referring to the original purchase — otherwise GA4 can subtract a phantom amount.
Partial refunds specify the refunded amount, not the full transaction value.

Verification SQL (BigQuery)

SELECT
  event_date,
  SUM(ecommerce.purchase_revenue) AS ga4_revenue,
  COUNT(*) AS purchase_events,
  COUNT(DISTINCT ecommerce.transaction_id) AS unique_transactions,
  COUNT(CASE WHEN ecommerce.purchase_revenue IS NULL THEN 1 END) AS null_revenue_rows
FROM `your-project.analytics_XXXXXXX.events_*`
WHERE event_name = 'purchase'
  AND _TABLE_SUFFIX BETWEEN '20260401' AND '20260420'
GROUP BY event_date
ORDER BY event_date;

If null_revenue_rows > 0, you have causes 1, 2, or 3 active. If purchase_events > unique_transactions, you have duplication (cause 6). Reconcile SUM(ga4_revenue) against your payment processor for the same period — a 2-5% gap is expected (timing + refunds); 10%+ warrants digging.

Duplicate Events in GA4 The dedup behaviour described in cause 6 explained in full.

Duplicate Transaction Prevention Prevent the duplication in the first place with a transaction-ID cache.

Data Discrepancies Expected vs. unexpected differences between GA4 and payment processor revenue.

BigQuery Schema Reference Where to find ecommerce fields in the BigQuery export for audit queries.

Missing Revenue in GA4

1. value parameter missing or not mapped

2. value is a string instead of a number

3. Missing or wrong currency code

4. items array missing prices