Short version: Broken conversion tracking has three symptom families, and each one has a different set of causes. Zero conversions means the tag never fired, the conversion action is misconfigured, or you are simply looking too early and attribution lag is doing the rest. Numbers that look too low against your platform usually come down to consent mode, browser tracking prevention, and the fact that Google Ads and GA4 measure attribution differently and will never match exactly. Numbers that look too high are almost always duplication: a tag firing twice, a lead action counting every conversion instead of one, or a purchase event that refires when someone refreshes the thank-you page. This guide walks the diagnostic tree in order, then covers how to verify the fix and what to do about the period of bad data.
The reason this deserves a methodical walkthrough rather than a quick checklist is that the wrong diagnosis wastes weeks. A store that "fixes" a consent problem when the real issue is a duplicate tag ends up with numbers that are still wrong, just wrong in a new direction, and no clearer idea of which report to believe.
Why broken tracking is the most expensive silent failure
A paused campaign announces itself. Broken tracking does not. Spend continues, clicks continue, the account looks alive, and meanwhile every decision made on top of the data is being made on fiction. The direct cost is bad reporting; the compounding cost is what smart bidding does with it.
Target ROAS and Maximize Conversions bid against the conversion data you feed them. Feed them nothing and they bid blind, usually retreating to cheap, low-intent traffic. Feed them doubled purchases and they scale aggressively into performance that does not exist. Either way the algorithm is not malfunctioning; it is faithfully optimizing toward a lie. This is why tracking sits ahead of every other item on our audit checklist: budget changes, bid changes, and structure changes are all downstream of whether the conversion column can be trusted.
Symptom one: zero conversions recorded
Total silence is actually the easiest symptom to diagnose, because the cause list is short and the failure is binary. Work through it in this order.
The tag never fires
The most common cause is the simplest: the conversion tag is not on the page, or not attached to the purchase event, so there is nothing to record. This happens more often than anyone admits, and rarely because someone forgot at setup. Checkout platform updates are the classic trigger. A theme change, a checkout migration, or a platform release quietly removes or breaks the snippet that was working last month, and nobody notices because nothing visibly changed on the site. If your conversions stopped on a specific date, the first question is what shipped on the site that day.
Consent mode is the second version of the same failure. If your consent banner blocks tags until the visitor accepts, and the configuration is wrong, the conversion tag can be blocked for everyone rather than just for the visitors who declined. The tag exists, the code is correct, and it still never runs.
The conversion action is misconfigured
Sometimes the tag fires and Google still records nothing usable, because the conversion action itself is set up wrong. A conversion action left as secondary rather than primary reports conversions but does not feed them to bidding. An action pointed at the wrong tag ID, or a Google Analytics import that was never completed, produces the same silence. Open the conversion action in Google Ads and read its status line before assuming the site is at fault; the diagnostics there will tell you whether Google has seen the tag recently at all.
It might just be lag
Before rebuilding anything, check the calendar. Conversions are reported against the date of the click, not the date of the purchase, and they arrive with delay. A conversion recorded today for a click last Tuesday appears in last Tuesday's row. Looking at today's numbers at lunchtime and concluding tracking is broken is the most common false alarm in Google Ads. If spend is recent, wait a full day or two and compare a completed date range before declaring an emergency.
Symptom two: the numbers are too low
This is the murkiest family, because some of the gap is a genuine problem and some of it is measurement physics that no fix will remove. The skill is telling those apart.
Consent and browser losses are real, and partly unavoidable
Visitors who decline cookies are not tracked directly, and iOS and browser tracking prevention trims more on top. Depending on your market and your banner design, that is a meaningful slice of real conversions that Google Ads will never see at full fidelity. Consent mode with modeling recovers a portion of it statistically, which is worth setting up properly, but the honest position is that a single-digit to moderate percentage gap against your platform is normal in a consented world. What is not normal is a gap of half your orders, which points back to a firing problem rather than a privacy one.
Google Ads, GA4, and Shopify will never match, by design
The comparison that generates the most panic is also the least meaningful. Google Ads uses click-based attribution: a conversion belongs to the ad click that led to it, credited on the click date, within a conversion window you chose. GA4 and most store platforms use session-based, last-touch logic on the purchase date. The same customer who clicked an ad on Monday, returned via email on Thursday, and bought, is a Google Ads conversion on Monday and an email conversion on Thursday in GA4. Neither report is lying; they are answering different questions. Expect direction and magnitude to agree over a month. Do not expect the totals to match on any given day, because the two systems are not counting the same thing.
The practical rule: judge Google Ads against itself over time, judge the business on blended platform revenue, and treat a sudden widening of the usual gap, rather than the existence of a gap, as the signal worth investigating.
Symptom three: the numbers are too high
Inflated numbers are more dangerous than missing ones, because they look like success. Nobody audits a winning account. The causes are almost always some form of duplication.
Duplicate tags firing on the same event
The classic setup accident: a conversion tag installed directly on the site and again through Google Tag Manager, or an old snippet left behind after a migration alongside the new one. Both fire on every purchase, and every order counts twice. If your Google Ads conversions run at roughly double your platform orders, this is the first place to look.
Counting "every" instead of "one"
Every conversion action has a counting setting. "Every" is correct for purchases, where three orders are three conversions. "One" is correct for leads, where the same person submitting a form four times is still one lead. A lead action left on "every" inflates conversion counts with repeat submissions, and smart bidding happily optimizes toward the people most likely to submit forms repeatedly, which is not the same audience as your best customers.
Purchase events that refire on refresh
If the purchase event fires on page load of the thank-you page rather than on the transaction itself, every refresh, back-button visit, or bookmarked return to that page records a fresh conversion. Transaction IDs exist precisely to deduplicate this: pass one with each purchase and Google discards repeats. A purchase action without transaction IDs is an inflation incident waiting for a customer who likes refreshing.
The verification workflow
Whatever you found and fixed, do not trust it until you have watched it work. The verification loop is short and worth running end to end.
First, load your checkout with Tag Assistant connected and watch the tags fire in real time: which tag, on which event, how many times, with what values. One clean fire per purchase with the right value and a transaction ID is the goal. Second, place a real test conversion, a genuine low-value order if you can, and confirm it appears in Google Ads with the correct value, remembering it will land against the click date. Third, read the diagnostics tab on each conversion action, which reports whether Google has received recent traffic from the tag and flags common problems on its own. Finally, compare a controlled window: pick a completed week after the fix, put Google Ads conversions next to platform orders from ad traffic, and confirm the gap has returned to the normal, explainable range rather than zero.
The platform-specific version of this setup, event by event, is covered in our Shopify conversion tracking walkthrough, which is the companion piece to this diagnostic guide.
What to do about the period of bad data
Fixing the tag does not fix the history. The weeks of missing or doubled conversions are still inside your reports and, more importantly, inside smart bidding's learning window.
Three practical moves. Annotate the broken period, in your reporting and your own records, so nobody later reads a performance cliff or spike as a market event. Expect a relearning period after the fix: bidding that trained on broken data takes days to weeks to recalibrate once clean data flows, and performance during that stretch is not the verdict on the fix. And resist making strategic decisions, budget cuts, campaign pauses, target changes, based on any window that includes the broken stretch. Judge the account on clean data only, which is inconvenient and correct.
Frequently asked questions
Why does Google Ads show fewer conversions than Shopify?
Partly losses, partly definitions. Consent declines and browser tracking prevention remove a slice of trackable conversions, which is a real gap. The rest is definitional: Google Ads counts only ad-attributed conversions against click dates, while Shopify counts every order on the purchase date regardless of source. A modest, stable gap is normal. A sudden widening of that gap is the signal to investigate.
Why does Google Ads show more conversions than I actually got?
Almost always duplication: two tags firing on the same purchase, a lead action counting every submission instead of one per person, or a thank-you page event refiring on refresh because no transaction ID deduplicates it. Roughly double your real orders points at a duplicate tag; a smaller inflation points at refires or counting settings.
How long do conversions take to show up in Google Ads?
Typically several hours to a day, and they are reported against the date of the ad click rather than the purchase date. A sale today from a click last week appears in last week's numbers. Always compare completed date ranges, and treat any same-day reading as provisional before concluding tracking is broken.
Should I pause campaigns while tracking is broken?
Usually not, unless spend is high and completely blind. Pausing destroys momentum and delivers its own relearning cost when you resume. The better sequence for most accounts is to hold budgets steady, fix tracking as the top priority, avoid optimization decisions during the broken window, and let bidding relearn on clean data once it flows.
Bringing it together
Diagnosis in order beats fixing at random: zero conversions means the tag, the action configuration, or simple lag; low numbers mean consent losses plus attribution definitions that guarantee Google Ads and your platform will never match exactly; high numbers mean duplication in one of its three usual forms. Verify with Tag Assistant, a test conversion, the diagnostics tab, and a controlled comparison window, then quarantine the bad-data period so it cannot contaminate decisions. Tracking is the foundation every other optimization stands on, which is why it leads our audit checklist and why the Shopify setup guide exists as its companion. If you would rather have a second pair of eyes confirm what your account is actually recording, request a free audit and tracking is the first thing we check.
About the author
This guide is written by Vasant Chaudhary, a Google Ads specialist with more than five years of experience managing over 50 ecommerce and lead generation accounts across the US, UK, and India. He focuses on conversion tracking, measurement, and campaign structure, the plumbing that decides whether an account's numbers can be trusted at all. Get in touch or start with a free audit call.
