Recharge subscriptions + TrackLayer: server-side LTV attribution

Shopify-native attribution is usually strong on the first order and weak on everything that happens after it. The moment a customer moves into Recharge, recurring charges, skipped cycles, pauses, reactivations, and churn often live outside the browser event path that most ad platforms were originally trained on. That means the media team keeps reporting efficient customer acquisition while the finance team sees a very different retention curve a month later.

TrackLayer closes that gap by treating Recharge as the revenue authority for subscription state changes. Instead of hoping a platform can infer lifetime value from the first checkout alone, TrackLayer receives the webhook, normalizes the payload, applies the value logic you actually want, and routes the result server-side. LTV then becomes a calculation based on contract behavior and cash events, not a guess hidden inside last-click ecommerce reporting.

The hard part is not getting Recharge data into TrackLayer. The hard part is deciding what value a paid platform should optimize against. Some brands only want to bid on the first order because that is the cleanest audited cash event. Others want modeled LTV because the first shipment loses money but subscription cohorts pay back over time. Others still want a recurring revenue proxy such as MRR so optimization favors customers who stick.

Pick one model per optimization lane. Mixing first-order value on Monday, 90-day LTV on Wednesday, and rebill revenue on Friday is how training data turns incoherent. TrackLayer can store all three values, but the destination payload should have one primary bidding value.

const purchaseValue = firstOrderNetRevenue;

const firstOrderNetRevenue =
  initialCharge - discounts - shippingSubsidy;

const purchaseValue =
  avgCycles90d * netRevenuePerCycle;

const avgCycles90d = 3.4;
const netRevenuePerCycle = 27;

const purchaseValue =
  monthlyCharge * retentionWeight;

const monthlyCharge = 39;
const retentionWeight = 1.6;

Recharge setups with prepaid plans, delayed first shipments, or trial offers should not collapse every stage into one purchase. Treat the trial start as a predictive signal and the first paid capture as the monetization event. That lets TrackLayer keep the customer journey intact while preventing free-start volume from inflating reported ROAS.

if (event === "subscription/created" && chargeToday === 0) {
  emit("trial_started", { subscription_id, customer_id, trial_end_at });
}

if (event === "subscription/processed" && amountCharged > 0) {
  emit("trial_converted", { subscription_id, customer_id, value: purchaseValue });
}

If paid media only sees starts and never sees cancellations, ROAS becomes structurally overstated. Recharge cancellation events are where TrackLayer can correct that. The conservative approach is to keep the original purchase event for auditability and send a separate cancellation adjustment event with a negative modeled value so downstream optimization and reporting understand that the expected revenue was lost.

if (event === "subscription/cancelled") {
  emit("subscription_cancelled", {
    subscription_id,
    customer_id,
    value: -projectedRemainingLtv,
    reason: cancel_reason,
    destination_event: "MetaCancellationAdjustment",
  });
}

Once Recharge lifecycle data is flowing, TrackLayer `/cohorts` becomes more useful than simple top-line renewal reporting. You can segment cohorts by acquisition source, first product, quiz answer cluster, discount depth, and subscription cadence, then compare realized rebills against the modeled value used for bidding. Accounts that skip, pause, or compress order intervals can be marked at risk before they cancel outright.

atRisk =
  pause_count > 0 ||
  skip_count >= 2 ||
  payment_failures >= 1 ||
  days_since_last_charge > expected_cycle_days + 7;