How LeakAudit calculates its numbers

LeakAudit runs four independent checks against your store's real Shopify data every time you scan. Every number below comes from a documented, fixed formula — never a guess, and never anything made up to look impressive. Here's exactly how each one works, including the assumptions baked in, so you can judge for yourself how much to trust it.

1. Payment & FX Fee Drag

We look at your last 30 days of orders and add up the ones placed in a currency different from your store's own (a "foreign presentment currency"). We estimate a 1.8% conversion drag on that revenue — a conservative industry rule of thumb for the combined cost of currency conversion and cross-border card fees. That's it: cross-border revenue × 1.8% = the monthly estimate you see. We don't yet read your actual payment gateway's fee schedule, so treat this as a directional estimate, not an invoice-accurate figure.

2. Leftover App Script Bloat

We read your live theme's code and check it against a short list of known snippets left behind by popular apps (review widgets, popups, tracking scripts, and similar) after a merchant uninstalls the app but the leftover code stays in the theme. Each match found is estimated at a flat $45/month — a placeholder cost representing typical page-speed and conversion drag from unused third-party scripts, not a measurement of your specific traffic yet.

3. Negative-Margin SKUs

For every in-stock product variant that has a cost-per-item set, we calculate: price − cost − your shipping assumption. If that's below your target margin percentage (both configurable in Settings — default $5.00 shipping, 20% target margin), the SKU is flagged. The monthly impact assumes roughly 3 units sold per month for each flagged SKU — a placeholder until real per-SKU sales velocity is wired in, so treat the dollar figure as directional and the flagged SKU list itself as the reliable part.

4. Return Rate Drift

We look at your last 60 days of orders and calculate what share of them received any refund. We compare that to an 8% benchmark (a commonly cited average return rate for general e-commerce). If your rate is meaningfully above that, we estimate the monthly cost from your actual refunded amounts, scaled to a 30-day figure. We need at least 5 orders in the window to compute a rate we're willing to show you — fewer than that, and we tell you honestly that there isn't enough data yet rather than showing a misleading percentage.

Your Health Score

Starts at 100. Each check that's actively "Leaking" costs 15 points; each check that hit an error (couldn't complete, usually a permissions or connectivity issue on our end) costs 5 points. A check that just needs more setup or data doesn't cost you anything — we don't penalize you for us not having enough information yet.

What we don't do

We never estimate a number we can't trace back to a formula like the ones above. We don't use AI or machine-learning "black box" predictions for any of these figures. Every check either has enough real data to compute a real number, or it tells you plainly that it doesn't yet.