Installed a tracking app and the numbers still don't match?
Tracking apps are built to apply the same fix to every store — a clean purchase event, a server-side layer, narrowing the everyday gap. They're good at that. What they structurally can't do is diagnose something specific to your setup: a tag that silently stopped firing, a phantom duplicate GA4 property, revenue read from the wrong field. If your numbers still don't reconcile after installing one, that's usually exactly the kind of issue an audit is built to find — not a sign the app failed at its job.
This isn't an argument against tracking apps — most do exactly what they're designed to do. It's about understanding the boundary of what an automated, one-size-fits-all fix can actually catch, so you don't keep troubleshooting in the wrong place.
What an app is built to fix
The everyday, structural gap
Apps are generally effective at recovering some of what ad blockers and Safari ITP take — typically through a server-side layer — narrowing the normal 10–30% gap down toward single digits for a lot of stores.
A standard, clean purchase event
Most apps ship a well-formed purchase event out of the box, which is genuinely useful if your previous setup had none, or had one configured incorrectly from the start.
What an app structurally can't catch
A bespoke, store-specific failure
A missing destination mapping in your Google Ads conversion action, a duplicate GA4 property nobody remembers creating, a GTM container that stopped loading on checkout after a platform update — none of these are things a generic app scans for, because they're specific to how your particular setup evolved.
Revenue read from the wrong field
An app can pass revenue through correctly — but if the underlying data itself is being read from the wrong currency field upstream, or includes tax and shipping it shouldn't, the app will faithfully report a wrong number with full confidence.
Architecture-level conflicts
Two tracking systems fighting over the same conversion event, or a legacy tag still live alongside a new app's implementation, isn't something an app can see — it only knows about its own piece of the puzzle, not what else is running alongside it.
"An app installs the same fix on every store. It can't see the specific thing wrong with yours."
What to check before assuming the app failed
- →Run a real test purchase and trace it end to end — don't just compare dashboard totals after the fact.
- →Check for leftover tags or properties from before the app was installed — old setups rarely get fully removed.
- →Confirm the app's reported revenue against Shopify's actual order total for a small, specific date range.
- →If the gap is large (40%+) rather than the normal 10–30%, suspect a specific fault, not just "the app needs tuning."
A recent audit found exactly this pattern: a brand's tracking had already passed through tools and freelancers, but two GA4 properties were still double-counting and the main production tag wasn't firing — issues no automated scan would have caught. Read the full case →