Is optimising for AddToCart really better than Purchase?

On low volume, usually yes. A stable ad set optimising for a proxy event outperforms an unstable one optimising for the right event but never leaving learning.

How wide should my attribution window be?

Match it to your measured time-to-purchase. If half your orders come more than a week after first touch, a 7-day window is discarding half your attribution.

Pixel tracking for high-ticket Shopify stores

Updated 1 September 2026

A store selling ten two-thousand-dollar items a week has the same revenue as one selling four hundred fifty-dollar items, and a completely different tracking problem. Meta’s delivery system needs event volume to stabilise, and high-ticket stores do not produce it. Optimising against Purchase on low volume leaves ad sets stuck in learning indefinitely, spending erratically and expensively.

What changes on a high-ticket store

Pixel tracking differences on a high-ticket Shopify store
What	How it differs
Optimisation target	Purchase volume is too low. Optimise against InitiateCheckout or AddToCart instead.
Value	Accurate per-order value matters more than anywhere else — it drives value-based lookalikes.
Attribution window	Consideration runs weeks. Short windows attribute almost nothing.
AddPaymentInfo	Worth assigning as a diagnostic — every abandonment at this step is expensive.

Low volume and the learning phase

Meta wants roughly fifty conversion events per ad set per week before delivery stabilises. A store selling a two-thousand-dollar product at ten orders a week cannot produce that from Purchase alone, so the ad set never exits learning and delivery stays erratic and expensive.

The standard answer is to optimise against an event further up the funnel — InitiateCheckout, or AddToCart on very low volumes — where there is enough weekly signal to stabilise. You give up some precision and buy consistency, which on a high-ticket store is usually the better trade. Purchase still fires and still reports; it simply is not the optimisation target.

Long consideration periods and window selection

Nobody buys a two-thousand-dollar item on the first visit. The research period runs days to weeks, across devices, often ending in a direct visit typed into the address bar rather than another ad click. A 7-day window will attribute almost none of this to the ad that actually started it.

Value-based audiences matter more here than almost anywhere else, because the difference between customers is enormous. A pixel sending Purchase with an accurate value lets Meta build lookalikes weighted toward high-value buyers rather than treating a hundred-dollar accessory sale and a two-thousand-dollar system sale as the same event.

Setup checklist

1Count your weekly Purchase events before choosing an optimisation target — under 50, move up the funnel.
2Send an accurate value on every conversion event, not just Purchase.
3Widen attribution windows to match your real consideration period.
4Assign AddPaymentInfo so late-stage abandonment is visible.
5Build value-based lookalikes rather than flat purchase lookalikes.

Questions

Is optimising for AddToCart really better than Purchase?: On low volume, usually yes. A stable ad set optimising for a proxy event outperforms an unstable one optimising for the right event but never leaving learning.
How wide should my attribution window be?: Match it to your measured time-to-purchase. If half your orders come more than a week after first touch, a 7-day window is discarding half your attribution.