First order at a loss? Judge it on payback, not ROAS

In Noah Haynes's teardown of Everyday Dose, the first offer is built to break even or lose money, so first-order ROAS is supposed to look bad, and it cannot tell you whether the offer works. Payback can: the month a group of new subscribers has earned back what it cost to win them, first-order loss included.

The video, published on YouTube on 22 September 2026, walks through the offer: a deep discount and a stack of free gifts on the first order, with a monthly subscription required to get them. In the video's own arithmetic, a $49 first order turns into $792 over a year, which is why the brand can afford to make nothing on the order that gets someone in the door (video). That $792 is revenue, not profit, and it assumes the subscriber stays all year.

The same video names the weak point. Without a strong retention system, it says, most customers cancel before the second or third charge, and it calls this one of the biggest cancellation spikes you will see. It also says the brand knows its unit economics well enough to spend whatever the ads need. From outside, nobody can check that. From inside, it is a table you can build this week.

Why first-order ROAS is the wrong number here

A loss leader is a deliberate trade: a loss today for charges later. First-order ROAS only sees the loss. That is not a flaw in your campaign. It is a flaw in the number.

What to look at instead, in this order:

1. The first order's own margin. Put the real first order into the break-even ROAS calculator: the price after the discount, the cost of the goods and the gifts, shipping and fees. If the order loses money before any ad runs, the calculator tells you no ROAS breaks even. For a loss leader, that is the right answer.
2. The loss as acquisition cost. Add the first-order loss to what you paid in ads per new subscriber. That total is what it really costs to win one, your true customer acquisition cost, and it is what payback has to cover.
3. The cancellation spike. Count how many of each month's new subscribers are still active at the second charge, the third and the sixth. Everyone who cancels before the second charge takes the whole first-order loss with them and pays none of it back.
4. The channel that brought them. Split those cohorts by the channel of the first order. A channel that brings cheap first orders from people who cancel before the second charge is buying the loss without the payback.

Meta optimises to the first purchase event and credits itself for every one, so a subscriber who cancels before the second charge is as good a result to Meta as one who stays all year.

How to see your payback this week

Build the cohort table. Export your subscriptions with start date, cancellation date and the number of charges so far. Group them by start month, and count who is still active at each charge. The share that cancels each month is your churn rate, and the first months' churn is the spike.

Mark the first order's channel. Renewal charges run in your subscription app without a website visit, so unless someone sends them to GA4 on purpose, GA4 sees the first order and not the renewals. Take the channel from the first order instead, from the UTM tags or the referrer your store recorded for the visit that led to it.

Work out payback per channel. For each channel's cohort, add up gross profit month by month from the charges that actually happened, first-order loss included, and find the month the running total passes what you paid in ads. That month is payback. The LTV to CAC calculator does a yearly version of the same sum: enter the renewal order's value and margin, count only the renewals as orders, and add the first-order loss to the cost to acquire a new customer, so the loss is counted once, as the acquisition cost it is.

Test the channel before you scale it. Switch one paid channel off in a few regions and compare new subscriptions by region from your store, then follow those subscribers to the second charge. The holdout test planner tells you how many days the regional test needs.

Where a read fits

Later, once the cohorts exist, a causal attribution read like Causality Engine's covers the other half. It takes the GA4 Attribution paths export and shows what each channel caused next to what last-click gave it, with Direct split back to the channels that sent those buyers. It sees the orders GA4 recorded, not the renewals your subscription app charged, so it tells you which channels cause orders and your cohorts tell you which of those buyers stay. Every channel gets a data-health score from 0 to 100 and a next step. It takes 1 to 2 minutes and costs EUR 99 once per upload, excluding VAT, with a full refund within 30 days, no questions asked. The first finding is free: your browser works out how many days buyers who saw two or more channels take to convert, and the file never leaves your machine.

Source, accessed 28 September 2026: Noah Haynes's teardown of Everyday Dose (YouTube, 22 September 2026).

Related answers

• Black Friday Discount Depth: The Loss You Book in Advance
• Is your welcome popup discount incremental?
• Does your customer community cause retention?
• The five-minute story ad that never gets the last click
• Hundreds of active ads: which angle caused the sales?
• Offer, creative, pages or retention: which drives growth?