How to Track Ecommerce Ad Performance

How to Track Ecommerce Ad Performance

1. 1

   Fire a purchase event that carries a value

   The single most important thing to get right is a purchase event that reports currency and value on every completed order, not just a page view or a generic conversion. Place a real test order and confirm the amount that arrives in the platform matches what the shopper paid. Without a value attached, the platform cannot optimise toward revenue and every report downstream is a count of orders with no sense of whether they were worth having.

2. 2

   Add the Conversions API alongside the pixel

   Browser pixels miss purchases when a shopper blocks tracking or the tab closes early, and that loss is uneven across devices. Sending the same purchase server side through the Conversions API, with proper deduplication so one order is not counted twice, recovers events the pixel drops. Follow the consent rules that apply to you. This does not make the platform figure exact, but it makes it less wrong and more stable week to week.

3. 3

   Keep order data in your own store admin

   Your store admin is the source of truth for money that actually arrived, because it records the order regardless of which ad, browser or attribution window was involved. Make sure it captures order value, date and, where you can, how the customer arrived. This is the number you reconcile everything against, and the only one that reflects refunds, cancellations and failed payments rather than an estimate made at the moment of checkout.

4. 4

   Reconcile the platform figure against your admin total

   Pick a period and compare platform reported revenue against the revenue your admin recorded for the same dates. They will not match. The platform credits sales inside its attribution window and adds modelled conversions for events it could not see, while your admin counts settled orders. Treat the platform number as directional, useful for comparing one ad against another, and trust your admin for what the business actually earned. If the gap is stable, you can learn to read across it.

5. 5

   Do not compare return across platforms that claim the same sale

   When a shopper sees an ad on one platform and another on a second before buying once, each platform can report that single sale as its own. Add the two returns together and you have counted the order twice, which makes blended performance look better than the money in your admin. Comparing return between platforms on their own reported figures is misleading for this reason. Judge each against the incremental sales you can see in your own totals, not against each other's claims.

6. 6

   Choose one number to check every week

   A dashboard with forty metrics is a way to avoid a decision. For a store owner the useful weekly number is usually total revenue attributed to ads set beside the revenue in your admin, or a blended return you calculate from your own sales rather than from any single platform. Pick the one number that tells you whether ad spend is producing profitable orders, check it on the same day each week, and record the offer that was running so the figure stays interpretable.

7. 7

   Read average order value and refunds, not just order count

   Two campaigns can produce the same number of purchases and very different money if one attracts a $34 basket and the other a $72 basket. Track average order value alongside the count, and check refunds and cancellations in your admin after the fact, because a purchase counted at checkout that is returned a week later was never revenue. The platform rarely sees the refund, so this correction only exists in your own numbers.