A monitoring tool is worth exactly as much as it is trusted. So this page explains the method rather than restating the claim, including the places the method stops.
A failed conversion looks exactly like a slow day
Conversion tracking runs in the customer's browser. It can be blocked by an extension, stopped by a consent banner, broken by a theme update, or simply never fire because the customer closed the tab a moment early.
When that happens nothing raises an error. Shopify records the order, because the order is real. Google Ads and Meta record nothing, because nobody told them. Neither system knows the other exists, so neither can notice they disagree — and the merchant sees a slightly quieter week.
Meanwhile the ad platforms keep optimising. Every bidding decision is made against the conversions they were told about, so a store losing a fifth of its tracking is buying traffic tuned to four fifths of its own customers.
Three numbers, one gap
- 1
Shopify tells us about every order
We subscribe to Shopify's order webhook, which fires on Shopify's servers the moment an order is created. No browser is involved, so nothing a customer's device does can suppress it. This is the number everything else is measured against.
Cancelled orders are excluded. Counting a refunded sale as a missed conversion would inflate the figure with sales that never happened.
- 2
Our pixel records what the browser saw
A Shopify web pixel extension runs at checkout and reports the order id it saw. It runs in the same conditions your Google and Meta tags do — same browser, same blockers, same consent rules — so when it fails, theirs almost certainly failed too.
It reports an order id and nothing else. No names, no email addresses, no basket contents.
- 3
The ad platforms tell us what they recordedComing next
For accounts you link yourself, we read the conversion counts Google Ads and Meta actually recorded over the same days. That closes the loop: it is possible for the browser to report a conversion and for the platform still not to have counted it.
Read-only, always. We never create, change or pause a campaign, a budget or a bid, and access can be revoked from your own account at any time.
- 4
The gap gets a price
Tell the app your monthly ad spend and it applies the loss rate to it: if a fifth of your conversions are missing, roughly a fifth of your spend is being optimised against incomplete data.
This is an estimate and is labelled as one everywhere it appears. It assumes the orders your browser failed to report look like the ones it did — the honest reading of the evidence, but an assumption, not a measurement.
Where these comparisons usually go wrong
Most of the work in a tool like this is not the comparison. It is the handful of decisions that make the comparison mean something.
- ·
Matching is on order id alone
Not on timestamps, which drift between a browser beacon and a webhook queue, and not on price, which is not unique. A report that arrives an hour late still matches its order.
- ·
Days are your store's days
Every figure is bucketed in your store's own timezone. Using the server's instead shifts every number by up to a day, which is enough to invent a problem or hide one.
- ·
A quiet day has no loss rate
A day with no orders reports nothing rather than zero. Zero is a claim that tracking was verified; the day was only quiet. Below ten orders no rate is published at all — one missed order in three is 33% and tells nobody anything.
- ·
Late reports are picked up
The trailing days are recomputed every hour, so a beacon that crossed midnight or arrived after a delay is counted rather than held against you.
What this cannot tell you
Every monitoring tool has an edge. These are ours, stated here rather than discovered later.
It cannot tell you which ad was missed
A missing conversion is missing — there is no record of where it came from. The gap is measured in orders and priced in spend, not attributed to a campaign.
The dollar figure is an estimate
It applies your loss rate to the ad spend you entered. If your untracked orders are systematically different from your tracked ones — cheaper, from a different channel — the estimate moves with them.
It does not fix the tracking for you
It tells you there is a gap, how large it is, and when it started. Repairing a tag, a theme or a consent banner is still work someone has to do.
A small store will see noise
Below roughly ten orders a day the rate swings on single events. That is why no rate is published under that threshold, and why the app is aimed at stores spending real money on ads.
One screen, and an email when it changes
The dashboard is the product. The alerts exist so you do not have to remember to open it.
Conversion data lost, last 30 days
18.2%
234 of 1,284 orders never reached your tracking.
See your own number
The free plan shows the loss rate and the orders behind it. That is enough to find out whether you have a problem.